favour in our resource conscious health service, particularly when an analysis of simple clinical factors reliably predicts the recurrence of seizures. The Royal College of Physicians' study showed that younger age, a family history of epilepsy or of febrile convulsions, and a seizure occurring between midnight and 8 59 am were factors associated with recurrence. If all three factors were present the recurrence rate at one year was 56% (95% confidence interval 42% to 70%); if two factors were present it was 43% (35% to 5 1%); if one factor was present 30% (25% to 36%); and if no factors were present 18% (7% to 28%). The study by Dr van Donselaar and colleagues also showed that younger age and seizures occurring at night or on waking were associated with recurrence. If these simple clinical questions can determine so well the probability of recurrence why bother with electroencephalography? The sensitivity of the first electroencephalogram in detecting epileptic discharges in patients after a first seizure in the study by Dr van Donselaar and colleagues was 16/151 (11%), and even if the recording was repeated after partial sleep deprivation (combined electroencephalograms) the number rose to only 35/151 (23%)-this with an interobserver agreement about epileptic discharges of only x=05. The authors were at pains to exclude all cases that were of doubtful idiopathic seizures. This low sensitivity (detection rate) of epileptic discharges in those who have had undoubted seizures should remind physicians that electroencephalography cannot be used to "diagnose" a doubtful event. The authors quote the specificity of combined electroencephalograms as being 91%-that is, a false positive rate of 9%. It is not clear what they mean by this as all patients in their study had truly had a first seizure. Neurologists traditionally interpret written electroencephalography reports in the light of the clinical history. Perhaps we should be more hard nosed. Clinical neurophysiologists would do us all a favour if they were more explicit about the sensitivity, specificity, and odds of being affected given a positive result5 for a variety of clinical disorders. Prior etalevaluated the use of sphenoidal electrodes in this way in 1975.6 Clinical neurophysiologists should surely build on this. ANTHONY HOPKINS Research Unit, Royal College of Physicians, London NWI 4LE I san Donselaar CA, Geerts AT, Schimsheimer R-J. Idiopathic first seizure in adult life: who should be treated? BMJ 1991;302:620-3. (16 March.) 2 Hauser WA, Rich SS, Annegers JF, Anderson ViE. Seizure recurrence after a first unprovoked seizure: an extended follow-up. Neurology 1990;40:1163-70. 3 Hopkins A, Garman A, Clarke CRA. The first seizure in adult life. Value of clinical features, electroencephalography and computerised tomographic scanning in prediction of seizure recurrence. Lancet 1988;i:721-6. 4 Salinsky M, Kanter R, Dashieff RM. Effectiveness of multiple EEGs in'supporting the diagnosis of epilepsy: an operational curve. Eptlepsia 1987;28:331-4. 5 Wald N. Rational use of investigations in clinical practice. In: Hopkins A, ed. Appropriate investigation anus treatment in clintcal practice. London: Royal College of Physicians, 1989. 6 Prior PF, Maynard DE, Scott DF. The value of sphenoidal EEG recording in patients with temporal lobe epilepsy. In: Epilepsologs. Proceedings of the seventh international svmposium on epilepss. Stuttgart: Georg Thieme, 1976.
SIR,-The well conducted hospital based study by Dr Cees A van- Donselaar and colleagues on the recurrence of a first idiopathic seizure' has replicated to some extent the results of a previous British survey.2 We would, however, make the following comment. The diagnosis of epileptic seizures is particularly difficult in its early stages. In 28% of patients enrolled in a community based study of epilepsy the diagnosis of the seizures was still unclear six months after entry.3 In the Mavo Clinic study, the time from the first seizure to diagnosis exceeded two years in more than 30% of patients.4 Dr van Donselaar and colleagues studied
only patients attending hospital with a well defined single attack, and the overall recurrence rate in this group (40%) may be different from the true rate of recurrence in less selected patients in the general population (78% in a recent population based study5). J W A S SANDER S D SHORVON National Hospital-Chalfont Centre for Epilepsy, Chalfont St Peter, Buckinghamshire SL9 ORJ I van Donselaar CA, Geerts AT, Schimsheimer R-J. Idiopathic first seizure in adult life: who should be treated? BMl7 1991;302:620-3. (16 March.) 2 Hopkins A, Garman A, Clarke C. The first seizure in adult life: value of clinical features, electroencephalography, and computerised tomographic scanning in prediction of seizure recurrence. Lancet 1988;i:721-6. 3 Sander JWAS, Hart YM, Johnson AL, Shorson SD. The national general practice study of epilepsy: newly diagnosed seizures in a general population. Lancet 1990;336:1267-71. 4 Hauser WA, Kurland LT. The epidemiology of epilepsy in Rochester, Minnesota 1935 through 1967. Epilepsia 1975;16: 1-66. S Hart YM, Sander JWAS, Johnson AL, Shorson SD. The national general practice study of epilepsy: recurrence after a first seizure. Lancet 1990;336:1271-4.
Ovarian hyperstimulation SIR,-In the editorial by Dr Blair H Smith and Professor Ian D Cooke the only preventive measure mentioned for the avoidance of the ovarian hyperstimulation syndrome in assisted reproduction was aspiration of follicles. In their letter in response to the article Messrs Martin S Mills and Peter G Wardle mentioned an alternative possibility of proceeding to oocyte retrieval and freezing all the oocytes with none being replaced.2 This avoids the chance of pregnancy, which can aggravate any hyperstimulation syndrome. There is an alternative approach that in our hands has proved highly effective. The principle involved is that in cycles controlled by gonadotrophin releasing hormone antagonists it is possible to withdraw gonadotrophin stimulation and have confidence that the surge of luteinising hormone will not occur. We adopted this approach after having three patients with serious hyperstimulation syndrome in 1987. Since mid-1988 we have applied this policy and have had no cases of the hyperstimulation syndrome in an in vitro fertilisation clinic treating more than 500 cycles a year. By maintaining the gonadotrophin releasing hormone antagonist down regulation and restimulating with gonadotrophins when the ovarian follicles have regressed, good pregnancy rates are obtained in such cycles.3 DAVID H BARLOW
DECLAN EGAN JULIAN ROBINSON Nuffield Department of Obstetrics and Gynaecology, John Radcliffe Hospital, Oxford OX3 9DU I Smith BH, Cooke ID. Ovarian hvperstimulation: actual and theoretical risks. BX,7 1991;302:127-8. (19 January.) 2 Mills MS, Wardle PG. Ovarian hyperstimulation. BMJ 1991; 302:474. (23 February.) 3 Forman R, Frydman R, Egan D, Ross C, Barlow D. Severe ovarian hvperstimulation syndrome associating LHRH agonists and gonadotrophins in IVF: a review of a European series and a proposal for prevention. Fertil Stentl 1990;53:502.
I desperately needed to see my son SIR,-MS Sheila Awooner-Renner reminds us all that grief is a very personal experience.' Guidelines are written in journals and in hospitals to assist in the care of the bereaved person. The main role of staff, however, must be to follow the relatives' expression of grief to be uninhibited by the pressures of social acceptance. There is no proper way
to grieve: in this tragic situation the response that feels right to the individual is right. Death is probably the last of the great taboos in Western society. People are scared to talk of the dead-their language becomes an entanglement of euphemisms, and the recently bereaved are socially isolated. In hospital many feel the need to protect recently bereaved people from "any more distress" by guiding them through the grief process. We should be accompanying them through this difficult journey but not deciding their route. No one should be afraid to ask them, "Is there anything you want us to do? Is there anything you want to do?" With a simple question we can dispel the agony of the hospital routine and leave them to cope with their loss as their only agony. Relatives' wishes are diverse-some may want to see a tidy relative in a shroud, others may prefer to see him as they did earlier that day, in his usual clothes and with his hair untidy as usual. Their last memory is their lasting memory. We must not let our fear and embarrassment add to their misery. Doctors are often too formal and aloof, either in reality or in perception. At King's College Hospital accident and emergency department we have recently employed a community link worker and one of her roles is the care of the bereaved. Because she is introduced as being there to help relatives she is seen as a friend rather than a figure of authority. She is not thought to have other work to rush away to. Relatives are more open with her and ask her questions more freely. Discussion with an apparent lay person is often less clinical but more informative at this stage. Clinical detail may be added by medical and nursing staff. Her continued contact over the next few weeks allows the opportunity to ask those often forgotten questions. Our link worker has received many letters and calls of thanks from bereaved relatives. The only complaint, from a young spouse, questioned the need for such vigorous resuscitation and her exclusion at the time of death-the decision of a doctor, not the link worker. MATTHEW W COOKE
Accident and Emergency Department, King's College Hospital, London SE5 9RS I Awooner-Renner S. I desperately needed to see my son. BMJ7 1991 ;302:356. (9 February.)
Triage of x ray films? SIR,-Drs I G H Renwick and colleagues have identified that the early reporting of radiographs obtained in accident and emergency x ray departments is becoming increasingly difficult.' Their results show that radiographers should not be expected to take on the task because of a considerable error rate. I agree that triage of x ray films "can be of use to casualty officers." At this hospital the department of radiology provides a "red dot" system, whereby radiographers triage x ray films before their interpretation by casualty officers. All x ray films are reported by a radiologist and important diagnostic errors are identified and rectified by our audit process. Over a recent 10 day period 352 radiographs were ordered by casualty officers, of which 40 showed abnormalities. Casualty officers had a false positive error rate of 1-7% of all films. Of the 40 abnormal films, eight were judged to be normal by both the radiographer and the casualty officer, giving a combined false negative error rate of 20% of the abnormal films, and 2 3% of all the films. These figures compare favourably with those in other studies.2' All abnormalities detected by radiographers were confirmed and properly treated by casualty officers. Further discussion showed that the presence of a red dot focused the casualty officers' attention on the presence of an abnormality and was thought
to be advantageous. Similarly, Montague and Glucksman found that casualty officers prescribed tetanus immunisation more appropriately if patients' immune state had been previously recorded by a triage nurse.4 Casualty officers seem to benefit from gentle reminders, especially when they are busy. I firmly believe that radiographers should be encouraged to triage accident and emergency radiographs; this has been shown to decrease the false negative error rate.' I K DUKES Accident and Emergency Department, Selly Oak Hospital, Birmingham B29 6JD 1 Renwick IGH, Butt WP, Steele B. How well can radiographers triage x ray films in accident and emergency departments? BMJ 1991;302:568-9. (9 March.) 2 Vincent CA, Driscoll PA, Audley RJ, Grant DS. Accuracy of detection of radiographic abnormalities by junior doctors. Arch EmergMed 1988;5:101-9. 3 Berman L, da Lacey G, Twomey E, Twomey B, Welch T, Eban R. Reducing errors in the accident and emergency department; a simple method using radiographers. BMJ 1985;290:421-2. 4 Montague A, Glucksman E. Influences on tetanus immunization in accident and emergency. Arch EmergMed 1990;7:163-8.
SIR,-Dr I G H Renwick and colleagues state that their objective was to assess radiographers' ability to identify abnormal radiographs.' This has already been done2; their study merely adds another category to the flagging system-namely, "insignificantly abnormal." This group was not defined. The importance of any abnormality is presumably a judgment for the clinician responsible for the patient. Therefore if the 58 radiographs in which radiographers in fact noted abnormalities but disagreed with the radiologists, saying that they were insignificant, are not included as false negatives, the false negative rate falls to 10% of abnormal radiographs and the overall error rate to 8%. Furthermore, Dr Renwick and colleagues made no assessment of the radiologists' accuracy. They state that they used the radiologists' reports as a gold standard, but in the next sentence refer to a paper which reports a 4% error rate and a 1% equivocal rate for radiologists reporting radiographs in accident and emergency departments.3 Indeed, Seltzer et al stated in their study that 6-3% of junior radiologists' reports were altered by senior colleagues.4 Radiographers' seniority has likewise been shown to affect their accuracy,2 but no assessment was made of this by the authors. This is particularly important as accident and emergency departments are commonly staffed with a large proportion of junior radiographers. Consequently, although we do not disagree wholeheartedly with the concluding statement of Dr Renwick and colleagues, we believe that their results suggest that there is a small total error rate ofdetection by radiographers, who are not specifically trained to diagnose radiographic abnormalities. Therefore there exists a considerable potential for radiographers to undergo a short period of training and provide an accurate means of identifying abnormal radiographs. Whether radiologists and radiographers will have the courage to attempt this and to evaluate the experience remains to be seen. H J NAWROCKA
Radiography Education Cenitre, Guy's Hospital, London J D NAWROCKI
l)epartment of Surgery King's Collcge Hospital, London I Renwick IGH, Butt WI', Steele B. How well can radiographers' triage x ray films in accident and emergency departments?
B.A11J 1991;302:568-9. 9,NMarch.) 2 Berman L, dc Lacey G, Twomev E, 'womey B, Welch T, Eban R. Reducing crrors in the accident and emergency department: a simple method tusing radiographers. BAIJ 1985;290:421-2. 3 De Lacey (iJ, Barker A, Harper J, Wignall B. An assessment of
the clinical efforts of reporting accident and emergency radiographs. BrJ Radiol 1980;53:304-9. 4 Seltzer SE, Hessel SJ, Herman PG, Swensson RG, Sheriff CR. Resident film interpretation and staff review. AIR 1981;137: 129-33.
War and medicine SIR, - I was concerned to read the letter on the war in Iraq from members of the North East Thames Medical Practitioners Union.' I think it is worth noting that most of the 100 000 dead they quote were members of the Iraqi armed forces and not civilians. It was obvious at the time of the conflict that the allied forces took great pains to avoid unnecessary civilian casualties. The Iraqi "compliance" with the United Nations resolution to withdraw from Kuwait was in fact a precipitate rout of an incompetently led and outclassed army. As soon as the political aims of the resolution had been achieved the allied advances ceased; if the allies had been out to inflict "gratuitous" casualties this would have occurred as the retreating Iraqi army withdrew into central Iraq. It has become increasingly obvious since Iraq's withdrawal from Kuwait that the Iraqi leadership has a total contempt for human life and also for the environment both locally within the Gulf and on a more global scale, with the firing of oil wells and the release of a large oil slick. Dr Anna Livingstone and colleagues would do better to devote their sympathies to the unfortunate Kurds who are being attacked by the helicopter gun ships of the Iraqi army, most of the victims in this case being unarmed civilians. Perhaps Iraq's oil revenues would be better spent improving the country's own humanitarian and medical infrastructure rather than on a genocidal war within its own boundaries.
is on record; and reports should be written at regular intervals by trainers, the accumulated information (facts about experience and opinions about performance) creating a much clearer picture of the candidate's true worth than the current selective and partial referehces. It is unfortunate that so much resistance (by both trainers and trainees) to these two potentially helpful reforms to our training arrangements still exists. JOHN R BENNETT
Gastrointestinal Unit, Hull Royal Infirmary, Hull HU3 2JZ I Arnold F. The research fetish. B.M7 1991;302:855. (6 April.)
ChanIges to thle NHS
SIR,-Considerable attention has been paid recently to the slowing of the reform of the NHS,' slowed further by the replacement of Mr Clarke with Mr Waldegrave at the Department of Health and then Mrs Thatcher with Mr Major as Prime Minister. There are, however, trends, identified with the white paper Working for Patients, that point to a change in attitudes to the NHS by policy makers and managers. Firstly, the white paper calls for explicit rationing of care (according to priorities) by purchasers to replace implicit or tacit rationing by providers after a process of free referral. Now we learn that the chief executive of the NHS Management Executive, Mr Duncan Nichol, has instructed regional health authorities to set conditions for access to waiting lists.2 However minor these may be in practice (for political reasons), an important precedent has been set in implementing the white paper's philosophy. T J JONES Secondly, the language of consumer choice is Department of Histopathology and Cytology, Royal Shrewsbury Hospital North, increasingly deployed, but in inverse proportion to Shrewsbury, the possibilities for choice by consumers and Shropshire SY3 8XQ patients since the implementation of the NHS and Community Care Act. This also shadows the white 1 Livingstone A, Patel RAJ, Lehmann AB. Pollen R, Owen A. paper's emphasis on working for patients in its War and medicine. BMJ 1991;302:849. (6 April.) rhetoric yet on making the manager (rather than the patient's advocate, the doctor) responsible for translating needs and priorities into contracts with providers. Yet a basic fact remains: in Britain The research fetish contracts made by purchasers with providers SIR,-Research should be performed and reported reflect the need for economy as a first priority. It is in the gradual abandonment of a radical for the right reasons, as Mr Frank Arnold cogently and wittily argued, and it is obviously undesirable market strategy, internal or otherwise, that if the selection of candidates by appointments undoubtedly the brakes have been applied. But the committees is based on the number of their use of the white paper as a tool for managers to publications and so encourages a valueless paper- enforce economy and reorient employee relations chase. A problem does exist and deserves public throughout the NHS, just as the competitive tendering process instigated in 1983 did in a debate; I should like to make two points. The precept that "research trains the mind" limited sphere, is increasingly developed. The purchaser-provider separation is a myth should not too readily be abandoned. Aspiring specialists who have never thought of a question except where NHS trusts are the providers-and, they would like to answer or a hypothesis they even there, Mr Waldegrave has recently announced should like to test are unlikely to become consult- significant restrictions on the freedom of trusts to ants who will properly assess the worth of new pay market rates and invest via capital markets. developments rather than simply follow fashion. For purchasers (districts) in fact employ and train Systematic audit is going to demand those the key workforce-doctors-employed by analytical faculties that are learnt as research providers. In practice, closely linked providing techniques. Unfortunately, too many trainees have and purchasing management teams will be an not attempted to devise a project but have asked expanded bureaucratic force against which other to be "given" some research-thus does dull, voices have less of a hearing. So, less change than uninspiring toil displace the alternating excitement expected in one direction but more in another. Markets require competition to provide and frustration that characterise original research. Secondly, appointments committees may use efficiency. If "unnecessary" investments, failures, publications as a measure of ability because they and closures are not permitted there will be no have so few other objective data by which to assess competitive market. That is no bad thing. So why candidates. More important considerations- not admit that strategic planning is not a discredited technical ability, judgment, consistency of per- adjunct of pre-1990 eastern Europe but the formance-can in most cases be estimated only essential tool of the modern general management from the candidate's own statements and perhaps of the NHS? Then we could allow provision of gleaned from selective references. Two fairly services to reflect agreement among doctors, simple measures could correct this and perhaps put managers, and patients' advocates within allowed the publication record where it belongs: all trainees budgets. Moreover, the government would not should keep log books so that their true experience have the worst of both worlds-opprobrium it no
BMJ VOLUME 302
27 APRIL 1991