many of whom would not risk epidural cannulation in a patient receiving standard unfractionated heparin. There is evidence that the period of greatest risk of thrombosis in hip replacement surgery is during surgery itself, when the femoral vein is subject to stress from the heat of polymerisation of bone cement and from the hip being maintained in a forced position in which the vein may be folded or kinked.' As the addition of the drug to the formulary was based on studies in which the first dose was given 12 hours preoperatively this cleared the way for enoxaparin to be added to the formulary. In Ireland the regulatory authorities have not stipulated that monitoring of antifactor Xa activity of enoxaparin is necessary, and the drug has been promoted on this basis. Furthermore, the recommended dose in joint replacement surgery is a standard 40 mg and is not based on weight. Indeed, the presentation of the drug in prefilled syringes discourages dose adjustment. Dose adjustment has not been shown to be necessary with enoxaparin and would reduce the simplicity of the drug's administration and hence the likelihood that this method of thromboprophylaxis would be adopted (A Planes, personal communication). Although based on a subjective measure-a considerable reduction in the incidence of clinically suspected deep vein thrombosis-this hospital's experience in the eight months since enoxaparin was introduced leads to the conclusion that our decision to use enoxaparin has been justified and that low molecular weight heparins will become the drugs of choice for this indication. TIM DELANEY
Pharmacy Department, Adelaide Hospital, Dublin 8, Republic of Ireland I Laserick MD, Croal SA, Mollan RAB. Orthopaedic surgeons and thromboprophylaxis. BMJ 1991;303:549-50. (7 Septem-
2 3 4
5
6 7
8
9
ber.) Fordyce MJF, Baker AS, Staddon GE. Efficacy of fixed minidose warfarin prophylaxis in joint replacement. BMJ 1991;303: 219-20. (27 July.) Hirsh J, Levine M. Prevention of venous thrombosis in patients undergoing major orthopaedic surgical procedures. Brj Clin Pract 1989;43 (suppl 65):2-8. Levvraz PF, Richard J, Bachmann F, Van Melle G, Treyraud JM, Livio JJ, et al. Adjusted versus fixed-dose heparin in the prevention of deep-vein thrombosis after total hip replacement. N EngljMed 1983;309:954-8. Leyvraz PF, Bachmann F, Hoek J, Buller HR, Postel Ai, Samama M, et al. Prevention of deep vein thrombosis after hip replacement: randomised comparison between unfractionated heparin and low molecular weight heparin. BMJ 1991;303: 543-8. (7 September.) Planes A, Vochelle N, Fagola M. Total hip replacement and deep vein thrombosis. A venographic and necropsy study. 7 Bone jointSurg[Br] 1990;72:9-13. Turpie AG, Levine MN, Hirsh J, Carter CJ, Jay RM, Powers PJ, et al. A randomized controlled trial of a low-molecular-weight heparin (enoxaparin) to prevent deep-vein thrombosis in patients undergoing elective hip surgery. N Engl J Med 1986;315:925-9. Frydman AM, Bara L, Le Roux Y, Woler M, Shauliac F, Samama MM. The antithrombotic activity and pharmacokinetics of enoxaparine, a low molecular weight heparin, in humans given single subcutaneous doses of 20 to 80 mg. J Clin Pharmacol 1988;28:609-18. Planes A, Vochelle N, Fagola M. rotal hip replacement and deep vein thrombosis. A venographic and necropsy study. J Bone Jooint Surg[Br] 1990;72:9-13.
SIR,-P F Leyvraz and colleagues' article on the use of fractionated heparin after hip replacement emphasises the reduction in the incidence of deep venous thrombosis after hip surgery.' Many other studies confirm this reduction, but the evidence that any form of prophylaxis with heparin reduces the incidence of fatal pulmonary embolism in hip surgery is scant,' reaching significance only if many series are summated.' In their editorial J Parker-Williams and R Vickers understate the case for warfarin,4 and the excellent results of Amstutz et al's regimen of low dose warfarin' have been completely overlooked in the articles' 46 and the subsequent correspondence. This is a remarkable omission in view of BMJ
VOLUME 303
9 NOVEMBER 1991
the reported zero incidence of fatal pulmonary embolism and fatal bleeding complications (and the 0-2% incidence of pulmonary embolism) in a series of 3000 patients compared with five pulmonary embolisms, one fatal pulmonary embolism, and one fatal bleeding complication in a series of only 349 patients.' The editorial's subtitle is "prophylaxis now or negligence claims later,"4 and the authors go on to advocate the use of an unproved agent, fractionated heparin. We emphasise Wroblewski and Triffit's views that this opinion is inappropriate.' Though we acknowledge that some form of prophylaxis is essential, potential complications such as wound haematomas in the treatment of hip fractures carry a high morbidity and mortality. Such complications from the use of prophylaxis of unproved efficacy could equally justifiably attract negligence claims. PETER W HOWARD RONAN B C TREACY PETER GRIGORIS
1 Parker-Williams J, Vickers R. Major orthopaedic surgery on the leg and thromboembolism. BMJ7 1991;303:531-2. (7 September.) 2 Aitkenhead A. Prudence with the pill. Journal of the Medical Defence Union 1990;winter:62-4.
Haemophilus influenzae type b invasive disease SIR,-A J Howard and colleagues report a high annual rate of infection with Haemophilus influenzae type b in children under the age of 5 years in Wales.' In Scotland 801 laboratory notifications of systemic infection with H influenzae (both type b and type not specified but presumably b) in children aged under 5 years were reported to the Communicable Diseases (Scotland) Unit by the bacteriology laboratories in Scotland during 1979-89 inclusive. The annual figures increased over the years (figure). In addition, 42 cases were
100
Royal Orthopaedic Hospital, Birmingham B31 2AP
901 Levvraz PF, Bachman F, Hoek J, Buller HR, Postel M, Samama M, et al. Prevention of deep venous thrombosis after hip replacement: randomised comparison between unfractionated heparin and low molecular weight heparin. BMJ 1991;303: 543-8. (7 September.) 2 Evarts CM, Alfadi RJ. Thromboembolism after total hip replacement. Failure of low dose heparin in prevention. 7AMA
1973;225:515-6. 3 Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. N Engl J Med 1988;318: 1162-73. 4 Parker-Williams J, Vickers R. Major orthopaedic surgery on the leg and thromboembolism. BM7 1991;303:531-2. (7
September.) 5 Amstutz HC, Friscia DA, Dorey F, Carnev BT. Warfarin prophylaxis to prevent pulmonary embolism after total hip replacement. 7 Bone Joint SurgfAm] 1989;71:321-6. 6 Laverick MD, Croal SA, Mollan RAB. Orthopaedic surgeons and thromboprophylaxis. BMJ 1991;303:549-50. (7 September.) 7 Wroblewski BM, Triffit PD. Orthopaedic surgeons and thromboprophylaxis. BMJ 1991;303:923. (12 October.)
SIR,-J Parker-Williams and Roger Vickers's alarmist editorial on thromboembolism and major orthopaedic surgery on the leg probably overstates the risk by suggesting that one in 40 patients die of pulmonary embolism after total hip replacement,' but nevertheless they have stimulated discussion and will challenge the 10% of surgeons who do not use any form of prophylaxis. Another controversy with regard to thromboembolism remains the use of the oestrogen contraceptive pill. We performed a brief postal survey of 35 major orthopaedic units in the United Kingdom and found a surprising variance of policies in our 26 replies. Seventeen of the units told patients to stop taking the pill: two specified that they should stop eight weeks before surgery, three that they should stop six weeks before, and 12 that they should stop four weeks before (though in three units this last applied to major surgery only). Six units allowed women to carry on using the pill; of these, three used heparin and three did not. Three units did not have a policy. The increased risk of venous thrombosis associated with the contraceptive pill is small, but even this small risk is diminished by stopping the pill or in emergencies by using heparin to increase antithrombin III activity, which is reduced by
O
80
I
t9
70 0 60-
z0
5040J.I
1979 1981 1983 1985 1987 1989 Number of infections with systemic H influenzae in children aged under 5 in Scotland, 1979-89
reported in children aged 5-9 years. There were 118 cases of epiglottitis and 618 of meningitis; "other infections" and cases in which blood cultures had been positive but there were no clinical details accounted for the remaining 107 cases. Of the children aged under 5 years, 76 were aged 0-5 months, 197 were aged 6-11 months, and 528 were aged 12-59 months. The population of children aged 0-59 months in any one year over this period was roughly 325 000; hence the overall annual notification rate was 22-4/100000 children under the age of 5. If, however, a particular cohort (children born in 1985) is studied the results are as shown in the table. The total number of children in the cohort was 68892, excluding 680 who died neonatally. The cumulative rate in this cohort was roughly 112/100 000; thus the chance of a child in Scotland being infected under the age of 5 years was one in 893, a much lower rate than reported by Howard and colleagues. Systemic infections with H influenzae in children born in 1985 Year of life
No
Rate/100 000
1 2 3 4 5
21 22 18 13 3
305 31-9 26-2 18-9 4.4
Total
77
112/100000
oestrogen.2
With patients' increasing awareness of the dangers of oral contraceptives containing oestrogen (however alarmist these may be), it is still cause for concern that roughly a fifth of orthopaedic units either have no policy or, perhaps worse, choose to ignore the risk. C B D LAVY T W R BRIGGS
Royal National Orthopaedic Hospital, Stanmore, Middlesex HA7 4LP
Inevitably there must be underascertainment of
invasive H influenzae type b infections as the
diagnosis can be made only in a laboratory and accurate ascertainment depends on all infections being seen by a practitioner, appropriate samples always being taken, and the infecting organism being recognised. There is no evidence that bacteriological services and notifications are other than satisfactory in Scotland (and surveillance of meningococcal disease supports this view), so we
1203
believe that there is a real difference between the situation in Wales and that in Scotland. The absence of secondary cases in Wales is reassuring, but we support the institution of prophylaxis for family contacts. R J FALLON Department of Laboratory Medicine, Ruchill Hospital, Glasgow G20 9NB D REID
Communicable Diseases (Scotland) Unit, Ruchill Hospital 1 Howard AJ, Dunkin KT, Musser JM, Palmer SR. Epidemiology of Haemophilus influenzae type b invasive disease in Wales. BMJ 1991 ;303:441-5. (24 August.)
Corticosteroids and tuberculosis SIR,-Martin B Allen and Nigel J Cooke's review of the role of corticosteroids in the management of tuberculosis omitted to mention two important practical considerations. ' Evidence suggests that adrenocortical function is compromised in patients with tuberculosis.' Rifampicin is a known inducer of the hepatic microsomal enzyme system and has been reported to cause enhanced clearance of endogenous cortisoll and to induce acute hypoadrenalism in tuberculous patients.4 Consideration of the patient's adrenal function and the requirement for steroid replacement may therefore need to be addressed as part of the treatment regimen. More importantly in a worldwide context is the requirement to exclude other potentially dangerous conditions that may be exacerbated by treatment with steroids. Steroids given to patients with intestinal strongyloidiasis may precipitate disseminated strongyloidiasis, which carries a high mortality.5 Similarly, amoebic disease has been reported to be precipitated by corticosteroids prescribed for tuberculous pleural effusions.6 Patients with tuberculosis who have lived in developing countries should be screened for these diseases before steroid treatment is started. BRYAN JEFFERSON HEAP Department of Public Health Medicine, St Andrew's Hospital, Norwich NR7 OSS 1 Allen MB, Cooke NJ. Corticosteroids and tuberculosis. BM7 1991;303:871-2. (12 October.) 2 Mugusi F, Swai ABM, Turner SJ, Alberti KGMM, McLarty DG. Hypoadrenalism in patients with pulmonary tuberculosis in Tanzania: an undiagnosed complication? Trans R Soc Trop Med Hvg 1990;84:849-51. 3 Edwards OM, Courtney-Evans RJ, Galley JM, Hunter J, Tait AD. Changes in cortisol metabolism following rifampicin therapy. Lancet 1974;ii:549-5 1. 4 Wilkins EGL, Hnizdo E, Cope A. Addisonian crisis induced by treatment with rifampicin. 7Tubercle 1989;70:69-73. 5 Stewart JB, Heap BJ. Fatal disseminated strongyloidiasis in an immuno-compromised former prisoner of the Japanese. J R Army Med Corps 1985;131:47-9. 6 Ratcliffe GE. Amoebic disease precipitated by corticosteroids prescribed for tuberculous pleural effusions. Tubercle 1988;69:2 19-2 1.
Radiographic examination of the lumbar spine SIR,-The message of Shawn F S Halpin and colleagues' paper on radiographic examination of the lumbar spine is clear. Except as a screen for spinal tumours, plain radiographs of the lumbar spine will not provide a doctor with a diagnosis for back pain when divorced from a good clinical examination. Even with a careful clinical assessment the diagnosis is often elusive, and it is never provided by plain radiographs alone. In fact, the situation is made worse: innocent degenerative change, or an incidental spondylolysis, will con-
1204
fuse the doctor and patient alike. For patients to be told that they have arthritis when the degenerative changes may be symptomless is counterproductive. Sometimes, however, a referral is to exclude tumour. Altogether 14% of patients in an orthopaedic clinic stated that the main reason for attending was "to be quite sure that there is nothing really serious wrong."2 Until we have noninvasive techniques readily available to exclude vertebral tumours patients and doctors will rightly request radiographs. Having excluded tumour, however, we should not read too much into isolated x ray films. In mechanical derangements of the lumbar spine, only after a good clinical assessment will an x ray film lend support to the diagnosis, as with the reduced lumbar lordosis due to protrusion of a disc, the degeneration of a disc causing root entrapment, the traction spurs of an unstable segment, or degenerative spondylolisthesis or structural scoliosis in neurogenic claudication. Even these changes need cautious interpretation, but if the clinician does not aim at a clinical diagnosis before radiography he or she is merely fishing in the dark, and radiography is unproductive. Only better education and dialogue will resolve the problem of radiography of the lumbar
spine. R W PORTER
elimination can be achieved, as with diphtheria and polio, if the vaccine is used correctly. It is appropriate that Ramsay and colleagues are examining the serum of vaccinated children for the type specific agglutinins 2 and 3. But even the traditional tests with unabsorbed and absorbed serum require caution to avoid pitfalls.7 Also, it is still too soon to rely on purified agglutinogens and enzyme linked immunosorbent assay (ELISA) for this purpose: vaccination with one such preparation, designated agglutinogen 3, produced antiserum that reacted in an ELISA not only with agglutinogen 3 but with agglutinogen 2 too,' thereby casting doubt on the purity of the agglutinogen and on the validity of those ELISAs for ensuring that antiserum contains both of the type specific agglutinins. NOEL W PRESTON
Pertussis Reference Laboratory, University Medical School, Manchester M13 9PT 1 Ramsav M, Begg N, Corbel MJ. Accelerated immunisation with diphtheria-tetanus-pertussis vaccine. BMJ 1991303:648-9. (14 September.) 2 Preston NW. Accelerated immunisation with diphtheria-tetanuspertussis vaccine. BMJ 1991;303:248. ('27 July.) 3 Andersen EK. Serological studies on H pertussis, H parapertussis and H bronchisepticus. Acta Pathologica et Microbiologica Scandinavica 1953;33:202-24. 4 Preston NW. Type-specific immunity against whooping cough.
BMJ17 1963;ii:724-6.
Department of Orthopaedics, Aberdeen Universitv Medical School, Aberdeen AB9 2ZD J A N SHEPPERD
Department of Orthopaedics, Royal East Sussex Hospital, Hastings, East Sussex TN34 1ER I Halpin SFS, Yeoman L, Dundas DD. Radiographic examination of the lumbar spine in a community hospital: an audit of current practice. BM7 1991;303:813-5. (5 October.) 2 Roland MO, Porter RW, Miatthews JG, Redden JF, Simonds GW, Bewley B. Improsing care: a study of orthopaedic outpatient referrals. Bill7 1991;302:1124-8.
Assessing the response to pertussis vaccine SIR,-Mary Ramsay and colleagues' note that my brief letter on the accelerated schedule for the primary course of three doses of diphtheriatetanus-pertussis vaccine did not mention boosters.2 My concern is that the primary response should be optimal: the weaker that response the greater will be the necessity for boosters. As yet, the need for boosters is neither proved nor disproved. Ramsay and colleagues' reference to "highly effective" pertussis vaccination "initiated at 1-2 months of age" relates to a time when type 1,2 strains were predominant (the reference was published in 1947). Andersen had not yet published her discovery of different serotypes'; and it was another 10 years before the first evidence of type specific immunity appeared.4 Today, type 1,3 strains are the main threat in many countries, and they produce disease in children with a deficiency of agglutinin 3. We must not revert to our position in the 1960s when vaccine deficient in agglutinogen 3 resulted in poor protection.' Moreover, in children who now receive vaccine containing all three agglutinogens the agglutinin 3 response is the weakest. At present it is not known whether vaccination before 3 months of age gives an adequate agglutinin 3 response. Ramsay and colleagues refer to the American schedule, which advises vaccination at 2, 4, and 6 months. This, however, does not constitute evidence of adequate response at those ages: enforcement of vaccination does not occur there until school entry, and vaccination at the recommended ages has been the exception rather than the rule." Moreover, we should not be content with the "substantial reduction in cases of pertussis and deaths from the disease in the US": virtual
5 Preston NW. Pertussis today. In: Wardlaw AC, Parton R, eds. Pathogenesis and immunity in pertussis. Chichester: Wiley, 1988:1-18. 6 Nkowane BM, Wassilak SGF, McKee PA, O'Mara DJ, Dellaportas G, Istre GR, et al. Pertussis epidemic in Oklahoma: difficulties in presenting transmission. Am J Dis Child 1986;140:433-7. 7 Preston NW, Surapatana N, Carter EJ. A reappraisal of serotvpe factors 4, 5 and 6 of Bordetella pertussis. journal off Hygiene
(Cambridge) 1982;88:39-46.
8 Ashworth LAE, Robinson A, Funnell S, Gorringe AR, Irons LI, Seabrook RN. Agglutinogens and fimbriae of bordetella pertussis. TokaiJ Exp Clin Med 1988;13(suppl):203- 10.
Budget holding: the first 150 days SIR, -It is a shame that John Bain chose to look at a budget holding practice that has obviously not performed as well as many in other areas. It \is with regret that I have to inform you that practdces in this area have not only increased surgical output for cold admissions over the first six months of fundholding compared with the previous 12 months but opened a ward with money from fundholding in a joint venture with the district health authority to benefit all our townsfolk; in addition, this fundholding practice has forced the local hospital to allow open access physiotherapy to all general practitioners, whether fundholders or not. These are positive moves, made by practices that wish to improve health care. D P M ARCHER Lincoln House Surgery, Hcmel Hempstead, Herttordshire HP2 4TU 1 Baini J. Budget holding: the first 150 days in Calserton. BMJ 1991;303:907-8. (12 Octoher.)
Training in the new NHS SIR,-John Bain's article on the first 150 days of budget holding by the Calverton practice highlights one of the effects that the "new NHS" will have on medical training.' The new dermatology service is provided by a consultant in a clinic in a private hospital. The rheumatology service is awaiting relocation. Where is the role for junior doctors and medical students in this system? Will consultants travel around the community holding clinics with a row of juniors and students perched on the back seat of their cars? Training centres rely on the centralisation
BMJ
VOLUME 303
9 NOVEMBER 1991
Pharmacy Department, Adelaide Hospital, Dublin 8, Republic of Ireland I Laserick MD, Croal SA, Mollan RAB. Orthopaedic surgeons and thromboprophylaxis. BMJ 1991;303:549-50. (7 Septem-
2 3 4
5
6 7
8
9
ber.) Fordyce MJF, Baker AS, Staddon GE. Efficacy of fixed minidose warfarin prophylaxis in joint replacement. BMJ 1991;303: 219-20. (27 July.) Hirsh J, Levine M. Prevention of venous thrombosis in patients undergoing major orthopaedic surgical procedures. Brj Clin Pract 1989;43 (suppl 65):2-8. Levvraz PF, Richard J, Bachmann F, Van Melle G, Treyraud JM, Livio JJ, et al. Adjusted versus fixed-dose heparin in the prevention of deep-vein thrombosis after total hip replacement. N EngljMed 1983;309:954-8. Leyvraz PF, Bachmann F, Hoek J, Buller HR, Postel Ai, Samama M, et al. Prevention of deep vein thrombosis after hip replacement: randomised comparison between unfractionated heparin and low molecular weight heparin. BMJ 1991;303: 543-8. (7 September.) Planes A, Vochelle N, Fagola M. Total hip replacement and deep vein thrombosis. A venographic and necropsy study. 7 Bone jointSurg[Br] 1990;72:9-13. Turpie AG, Levine MN, Hirsh J, Carter CJ, Jay RM, Powers PJ, et al. A randomized controlled trial of a low-molecular-weight heparin (enoxaparin) to prevent deep-vein thrombosis in patients undergoing elective hip surgery. N Engl J Med 1986;315:925-9. Frydman AM, Bara L, Le Roux Y, Woler M, Shauliac F, Samama MM. The antithrombotic activity and pharmacokinetics of enoxaparine, a low molecular weight heparin, in humans given single subcutaneous doses of 20 to 80 mg. J Clin Pharmacol 1988;28:609-18. Planes A, Vochelle N, Fagola M. rotal hip replacement and deep vein thrombosis. A venographic and necropsy study. J Bone Jooint Surg[Br] 1990;72:9-13.
SIR,-P F Leyvraz and colleagues' article on the use of fractionated heparin after hip replacement emphasises the reduction in the incidence of deep venous thrombosis after hip surgery.' Many other studies confirm this reduction, but the evidence that any form of prophylaxis with heparin reduces the incidence of fatal pulmonary embolism in hip surgery is scant,' reaching significance only if many series are summated.' In their editorial J Parker-Williams and R Vickers understate the case for warfarin,4 and the excellent results of Amstutz et al's regimen of low dose warfarin' have been completely overlooked in the articles' 46 and the subsequent correspondence. This is a remarkable omission in view of BMJ
VOLUME 303
9 NOVEMBER 1991
the reported zero incidence of fatal pulmonary embolism and fatal bleeding complications (and the 0-2% incidence of pulmonary embolism) in a series of 3000 patients compared with five pulmonary embolisms, one fatal pulmonary embolism, and one fatal bleeding complication in a series of only 349 patients.' The editorial's subtitle is "prophylaxis now or negligence claims later,"4 and the authors go on to advocate the use of an unproved agent, fractionated heparin. We emphasise Wroblewski and Triffit's views that this opinion is inappropriate.' Though we acknowledge that some form of prophylaxis is essential, potential complications such as wound haematomas in the treatment of hip fractures carry a high morbidity and mortality. Such complications from the use of prophylaxis of unproved efficacy could equally justifiably attract negligence claims. PETER W HOWARD RONAN B C TREACY PETER GRIGORIS
1 Parker-Williams J, Vickers R. Major orthopaedic surgery on the leg and thromboembolism. BMJ7 1991;303:531-2. (7 September.) 2 Aitkenhead A. Prudence with the pill. Journal of the Medical Defence Union 1990;winter:62-4.
Haemophilus influenzae type b invasive disease SIR,-A J Howard and colleagues report a high annual rate of infection with Haemophilus influenzae type b in children under the age of 5 years in Wales.' In Scotland 801 laboratory notifications of systemic infection with H influenzae (both type b and type not specified but presumably b) in children aged under 5 years were reported to the Communicable Diseases (Scotland) Unit by the bacteriology laboratories in Scotland during 1979-89 inclusive. The annual figures increased over the years (figure). In addition, 42 cases were
100
Royal Orthopaedic Hospital, Birmingham B31 2AP
901 Levvraz PF, Bachman F, Hoek J, Buller HR, Postel M, Samama M, et al. Prevention of deep venous thrombosis after hip replacement: randomised comparison between unfractionated heparin and low molecular weight heparin. BMJ 1991;303: 543-8. (7 September.) 2 Evarts CM, Alfadi RJ. Thromboembolism after total hip replacement. Failure of low dose heparin in prevention. 7AMA
1973;225:515-6. 3 Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. N Engl J Med 1988;318: 1162-73. 4 Parker-Williams J, Vickers R. Major orthopaedic surgery on the leg and thromboembolism. BM7 1991;303:531-2. (7
September.) 5 Amstutz HC, Friscia DA, Dorey F, Carnev BT. Warfarin prophylaxis to prevent pulmonary embolism after total hip replacement. 7 Bone Joint SurgfAm] 1989;71:321-6. 6 Laverick MD, Croal SA, Mollan RAB. Orthopaedic surgeons and thromboprophylaxis. BMJ 1991;303:549-50. (7 September.) 7 Wroblewski BM, Triffit PD. Orthopaedic surgeons and thromboprophylaxis. BMJ 1991;303:923. (12 October.)
SIR,-J Parker-Williams and Roger Vickers's alarmist editorial on thromboembolism and major orthopaedic surgery on the leg probably overstates the risk by suggesting that one in 40 patients die of pulmonary embolism after total hip replacement,' but nevertheless they have stimulated discussion and will challenge the 10% of surgeons who do not use any form of prophylaxis. Another controversy with regard to thromboembolism remains the use of the oestrogen contraceptive pill. We performed a brief postal survey of 35 major orthopaedic units in the United Kingdom and found a surprising variance of policies in our 26 replies. Seventeen of the units told patients to stop taking the pill: two specified that they should stop eight weeks before surgery, three that they should stop six weeks before, and 12 that they should stop four weeks before (though in three units this last applied to major surgery only). Six units allowed women to carry on using the pill; of these, three used heparin and three did not. Three units did not have a policy. The increased risk of venous thrombosis associated with the contraceptive pill is small, but even this small risk is diminished by stopping the pill or in emergencies by using heparin to increase antithrombin III activity, which is reduced by
O
80
I
t9
70 0 60-
z0
5040J.I
1979 1981 1983 1985 1987 1989 Number of infections with systemic H influenzae in children aged under 5 in Scotland, 1979-89
reported in children aged 5-9 years. There were 118 cases of epiglottitis and 618 of meningitis; "other infections" and cases in which blood cultures had been positive but there were no clinical details accounted for the remaining 107 cases. Of the children aged under 5 years, 76 were aged 0-5 months, 197 were aged 6-11 months, and 528 were aged 12-59 months. The population of children aged 0-59 months in any one year over this period was roughly 325 000; hence the overall annual notification rate was 22-4/100000 children under the age of 5. If, however, a particular cohort (children born in 1985) is studied the results are as shown in the table. The total number of children in the cohort was 68892, excluding 680 who died neonatally. The cumulative rate in this cohort was roughly 112/100 000; thus the chance of a child in Scotland being infected under the age of 5 years was one in 893, a much lower rate than reported by Howard and colleagues. Systemic infections with H influenzae in children born in 1985 Year of life
No
Rate/100 000
1 2 3 4 5
21 22 18 13 3
305 31-9 26-2 18-9 4.4
Total
77
112/100000
oestrogen.2
With patients' increasing awareness of the dangers of oral contraceptives containing oestrogen (however alarmist these may be), it is still cause for concern that roughly a fifth of orthopaedic units either have no policy or, perhaps worse, choose to ignore the risk. C B D LAVY T W R BRIGGS
Royal National Orthopaedic Hospital, Stanmore, Middlesex HA7 4LP
Inevitably there must be underascertainment of
invasive H influenzae type b infections as the
diagnosis can be made only in a laboratory and accurate ascertainment depends on all infections being seen by a practitioner, appropriate samples always being taken, and the infecting organism being recognised. There is no evidence that bacteriological services and notifications are other than satisfactory in Scotland (and surveillance of meningococcal disease supports this view), so we
1203
believe that there is a real difference between the situation in Wales and that in Scotland. The absence of secondary cases in Wales is reassuring, but we support the institution of prophylaxis for family contacts. R J FALLON Department of Laboratory Medicine, Ruchill Hospital, Glasgow G20 9NB D REID
Communicable Diseases (Scotland) Unit, Ruchill Hospital 1 Howard AJ, Dunkin KT, Musser JM, Palmer SR. Epidemiology of Haemophilus influenzae type b invasive disease in Wales. BMJ 1991 ;303:441-5. (24 August.)
Corticosteroids and tuberculosis SIR,-Martin B Allen and Nigel J Cooke's review of the role of corticosteroids in the management of tuberculosis omitted to mention two important practical considerations. ' Evidence suggests that adrenocortical function is compromised in patients with tuberculosis.' Rifampicin is a known inducer of the hepatic microsomal enzyme system and has been reported to cause enhanced clearance of endogenous cortisoll and to induce acute hypoadrenalism in tuberculous patients.4 Consideration of the patient's adrenal function and the requirement for steroid replacement may therefore need to be addressed as part of the treatment regimen. More importantly in a worldwide context is the requirement to exclude other potentially dangerous conditions that may be exacerbated by treatment with steroids. Steroids given to patients with intestinal strongyloidiasis may precipitate disseminated strongyloidiasis, which carries a high mortality.5 Similarly, amoebic disease has been reported to be precipitated by corticosteroids prescribed for tuberculous pleural effusions.6 Patients with tuberculosis who have lived in developing countries should be screened for these diseases before steroid treatment is started. BRYAN JEFFERSON HEAP Department of Public Health Medicine, St Andrew's Hospital, Norwich NR7 OSS 1 Allen MB, Cooke NJ. Corticosteroids and tuberculosis. BM7 1991;303:871-2. (12 October.) 2 Mugusi F, Swai ABM, Turner SJ, Alberti KGMM, McLarty DG. Hypoadrenalism in patients with pulmonary tuberculosis in Tanzania: an undiagnosed complication? Trans R Soc Trop Med Hvg 1990;84:849-51. 3 Edwards OM, Courtney-Evans RJ, Galley JM, Hunter J, Tait AD. Changes in cortisol metabolism following rifampicin therapy. Lancet 1974;ii:549-5 1. 4 Wilkins EGL, Hnizdo E, Cope A. Addisonian crisis induced by treatment with rifampicin. 7Tubercle 1989;70:69-73. 5 Stewart JB, Heap BJ. Fatal disseminated strongyloidiasis in an immuno-compromised former prisoner of the Japanese. J R Army Med Corps 1985;131:47-9. 6 Ratcliffe GE. Amoebic disease precipitated by corticosteroids prescribed for tuberculous pleural effusions. Tubercle 1988;69:2 19-2 1.
Radiographic examination of the lumbar spine SIR,-The message of Shawn F S Halpin and colleagues' paper on radiographic examination of the lumbar spine is clear. Except as a screen for spinal tumours, plain radiographs of the lumbar spine will not provide a doctor with a diagnosis for back pain when divorced from a good clinical examination. Even with a careful clinical assessment the diagnosis is often elusive, and it is never provided by plain radiographs alone. In fact, the situation is made worse: innocent degenerative change, or an incidental spondylolysis, will con-
1204
fuse the doctor and patient alike. For patients to be told that they have arthritis when the degenerative changes may be symptomless is counterproductive. Sometimes, however, a referral is to exclude tumour. Altogether 14% of patients in an orthopaedic clinic stated that the main reason for attending was "to be quite sure that there is nothing really serious wrong."2 Until we have noninvasive techniques readily available to exclude vertebral tumours patients and doctors will rightly request radiographs. Having excluded tumour, however, we should not read too much into isolated x ray films. In mechanical derangements of the lumbar spine, only after a good clinical assessment will an x ray film lend support to the diagnosis, as with the reduced lumbar lordosis due to protrusion of a disc, the degeneration of a disc causing root entrapment, the traction spurs of an unstable segment, or degenerative spondylolisthesis or structural scoliosis in neurogenic claudication. Even these changes need cautious interpretation, but if the clinician does not aim at a clinical diagnosis before radiography he or she is merely fishing in the dark, and radiography is unproductive. Only better education and dialogue will resolve the problem of radiography of the lumbar
spine. R W PORTER
elimination can be achieved, as with diphtheria and polio, if the vaccine is used correctly. It is appropriate that Ramsay and colleagues are examining the serum of vaccinated children for the type specific agglutinins 2 and 3. But even the traditional tests with unabsorbed and absorbed serum require caution to avoid pitfalls.7 Also, it is still too soon to rely on purified agglutinogens and enzyme linked immunosorbent assay (ELISA) for this purpose: vaccination with one such preparation, designated agglutinogen 3, produced antiserum that reacted in an ELISA not only with agglutinogen 3 but with agglutinogen 2 too,' thereby casting doubt on the purity of the agglutinogen and on the validity of those ELISAs for ensuring that antiserum contains both of the type specific agglutinins. NOEL W PRESTON
Pertussis Reference Laboratory, University Medical School, Manchester M13 9PT 1 Ramsav M, Begg N, Corbel MJ. Accelerated immunisation with diphtheria-tetanus-pertussis vaccine. BMJ 1991303:648-9. (14 September.) 2 Preston NW. Accelerated immunisation with diphtheria-tetanuspertussis vaccine. BMJ 1991;303:248. ('27 July.) 3 Andersen EK. Serological studies on H pertussis, H parapertussis and H bronchisepticus. Acta Pathologica et Microbiologica Scandinavica 1953;33:202-24. 4 Preston NW. Type-specific immunity against whooping cough.
BMJ17 1963;ii:724-6.
Department of Orthopaedics, Aberdeen Universitv Medical School, Aberdeen AB9 2ZD J A N SHEPPERD
Department of Orthopaedics, Royal East Sussex Hospital, Hastings, East Sussex TN34 1ER I Halpin SFS, Yeoman L, Dundas DD. Radiographic examination of the lumbar spine in a community hospital: an audit of current practice. BM7 1991;303:813-5. (5 October.) 2 Roland MO, Porter RW, Miatthews JG, Redden JF, Simonds GW, Bewley B. Improsing care: a study of orthopaedic outpatient referrals. Bill7 1991;302:1124-8.
Assessing the response to pertussis vaccine SIR,-Mary Ramsay and colleagues' note that my brief letter on the accelerated schedule for the primary course of three doses of diphtheriatetanus-pertussis vaccine did not mention boosters.2 My concern is that the primary response should be optimal: the weaker that response the greater will be the necessity for boosters. As yet, the need for boosters is neither proved nor disproved. Ramsay and colleagues' reference to "highly effective" pertussis vaccination "initiated at 1-2 months of age" relates to a time when type 1,2 strains were predominant (the reference was published in 1947). Andersen had not yet published her discovery of different serotypes'; and it was another 10 years before the first evidence of type specific immunity appeared.4 Today, type 1,3 strains are the main threat in many countries, and they produce disease in children with a deficiency of agglutinin 3. We must not revert to our position in the 1960s when vaccine deficient in agglutinogen 3 resulted in poor protection.' Moreover, in children who now receive vaccine containing all three agglutinogens the agglutinin 3 response is the weakest. At present it is not known whether vaccination before 3 months of age gives an adequate agglutinin 3 response. Ramsay and colleagues refer to the American schedule, which advises vaccination at 2, 4, and 6 months. This, however, does not constitute evidence of adequate response at those ages: enforcement of vaccination does not occur there until school entry, and vaccination at the recommended ages has been the exception rather than the rule." Moreover, we should not be content with the "substantial reduction in cases of pertussis and deaths from the disease in the US": virtual
5 Preston NW. Pertussis today. In: Wardlaw AC, Parton R, eds. Pathogenesis and immunity in pertussis. Chichester: Wiley, 1988:1-18. 6 Nkowane BM, Wassilak SGF, McKee PA, O'Mara DJ, Dellaportas G, Istre GR, et al. Pertussis epidemic in Oklahoma: difficulties in presenting transmission. Am J Dis Child 1986;140:433-7. 7 Preston NW, Surapatana N, Carter EJ. A reappraisal of serotvpe factors 4, 5 and 6 of Bordetella pertussis. journal off Hygiene
(Cambridge) 1982;88:39-46.
8 Ashworth LAE, Robinson A, Funnell S, Gorringe AR, Irons LI, Seabrook RN. Agglutinogens and fimbriae of bordetella pertussis. TokaiJ Exp Clin Med 1988;13(suppl):203- 10.
Budget holding: the first 150 days SIR, -It is a shame that John Bain chose to look at a budget holding practice that has obviously not performed as well as many in other areas. It \is with regret that I have to inform you that practdces in this area have not only increased surgical output for cold admissions over the first six months of fundholding compared with the previous 12 months but opened a ward with money from fundholding in a joint venture with the district health authority to benefit all our townsfolk; in addition, this fundholding practice has forced the local hospital to allow open access physiotherapy to all general practitioners, whether fundholders or not. These are positive moves, made by practices that wish to improve health care. D P M ARCHER Lincoln House Surgery, Hcmel Hempstead, Herttordshire HP2 4TU 1 Baini J. Budget holding: the first 150 days in Calserton. BMJ 1991;303:907-8. (12 Octoher.)
Training in the new NHS SIR,-John Bain's article on the first 150 days of budget holding by the Calverton practice highlights one of the effects that the "new NHS" will have on medical training.' The new dermatology service is provided by a consultant in a clinic in a private hospital. The rheumatology service is awaiting relocation. Where is the role for junior doctors and medical students in this system? Will consultants travel around the community holding clinics with a row of juniors and students perched on the back seat of their cars? Training centres rely on the centralisation
BMJ
VOLUME 303
9 NOVEMBER 1991
