miscarriage, of whom 12 said that they would like to have been shown the products of conception even if no fetus could be identified. The desire to see he fetus or products was not related to the patients' age, parity, marital status, gestation, or planning of the pregnancy. These results indicated that one third of patients would like to be offered the opportunity of seeing the fetus after a spontaneous miscarriage. Whether such a measure would help the patient to grieve remains to be determined. In the absence of scientific proof, however, there is a strong case for acceding to the wishes of the parents if they ask to see the fetus after a spontaneous miscarriage. Miscarriage is the commonest complication of pregnancy, but we still have much to learn about the best way to help couples to come to terms with their loss. MICHAEL J TURNER Coombe Lying-In Hospital, Dublin 8, Republic of Ireland 1 Awoonor-Renner S. I desperately needed to see my son. BMJ 1991;302:356. (9 Februars ..) 2 Forrest G. Care of the bereaved after perinatal death. In: Chalmers I, Enkin M, Keirse MJNC, eds. Effective care in pregnancy and childbinrh. Oxford: Oxford University Press, 1989:1423-30. 3 Iles S. The loss of early pregnancy. Clin Obstet Gynaecol 1989;3:769-90. 4 Turner MJ. Spontaneous miscarriage: this hidden loss. IrMedj 1989;89: 145. 5 Turner Mj, Flannellv GM, Wingfield M, et al. The miscarriage clinic: an audit of the first year. Br J Obstet Gvnaecol (in press).

Early pregnancy assessment SIR,-The article by Drs M A Bigrigg and M D Read exemplifies how by reorganising their obstetric services they could offer a better care and cost effective service for their patients with problems in early pregnancy.' The result of their study may also be well taken by the hospital managements who with the new trust arrangements would be happy to look into the cost effectiveness of any service offered under their trust care system. Drs Bigrigg and Read have omitted to mention who actually undertakes the ultrasonography on their patients-are their patients referred to the radiographers or is this service offered by the obstetricians? If it is done by the obstetricians it serves two purposes: diagnosis and further management counselling. It is heartening to report that North Devon District Hospital has been offering such care and cost effective treatment to its patients referred with bleeding or pain in early pregnancy, within limited medical staffing resources, and without the need for reorganisation of its gynaecology and obstetric services. The key point in being able to offer such a service by this unit is the fact that ultrasonography is undertaken by the consultants or registrar with the Toshiba sonolayer LSAL-77A with curvilinear abdominal transducers, which is installed in the maternity unit and is available for use any time of the day or night. This hospital serves a population of 150 700 and receives patients within a perimeter of about 72 km. The obstetric and gynaecological medical team consists of three consultants, one registrar, and three senior house officers. Women requiring assessment for bleeding or pain during early pregnancy, who are otherwise in satisfactory general condition, are accepted by the house officer on behalf of the consultant on call. Once the patient arrives at the hospital a brief history is taken, and a sample of venous blood is taken for full blood count and blood group analysis. Ultrasound scanning sessions are arranged between the morning and afternoon scheduled sessions, and the patients are scanned and counselled. Patients needing evacuation of the uterus are often dealt

910

with as an additional case added on to the end of the afternoon gynaecological operating list, and rarely some cases spill beyond the routine operating list. Other patients not requiring further treatment are discharged with further follow up appointments if necessary. This policy is also helpful in dealing with the extra load of work during the summer, when the district receives many holidaymakers and some of these women reach the hospital with early pregnancy complications. Although this practice of early assessment sometimes puts pressure on the consultant or registrar who undertakes the scanning procedure to decide further management, it is of greater benefit to the patient and to the hospital management. R KULKARNI W P BRADFORD

Department of Obstetrics and Gvnaecology, North Devon District Hospital, Bamstaple EX31 4JB 1 Bigrigg MA, Read MID. Management of women referred to earlv pregnancy assessment unit: care and cost effectiveness. B.MJ 1991;302:577-9. (9 MNarch.)

Epidural analgesia and maternal satisfaction SIR,-Dr J D Murphy and colleagues state that by adding fentanyl to conventional bupivacaine for epidural analgesia in labour it is possible to improve maternal satisfaction and increase the chances of a normal vaginal delivery.' Although the former assumption may be correct, their overall results suggest that a normal vaginal delivery with epidural analgesia is the exception rather than the rule. The results indicate an overall rate of normal vaginal deliveries of only 35% (30/85), a rate of caesarean sections of 20% (17/85), and a rate for a combined (simple plus rotational) forceps delivery of 40% (34/85). I suggest that this is not normal, particularly as the authors state in their methods that they were dealing solely with primiparous women at term with a singleton fetus and with a cephalic presentation. This category of women is the basis of the active management of labour,2 and the hospitals that adopt this policy in labour have remarkably constant figures for caesarean sections and forceps deliveries for their primigravid population. Indeed, the latest annual report from the maternity hospital in Dublin showed a rate of non-elective caesarean sections of 6% and of forceps deliveries of 11 9% in primigravidas. I realise that Dr Murphy and colleagues are anaesthetists and therefore are not responsible for the obstetric management of their delivery unit. None the less, I think that the reason for the overall low percentage of normal vaginal deliveries coupled with the high rates of caesarean section and forceps deliveries should have been stated in their paper. If these figures are par for the course in their delivery unit-and I would hope they are not-then better management of the mother's labour rather than her analgesia might give her greater satisfaction in the end. DECLAN P KEANE

Bristol Maternity Hospital, Bristol BS2 8EG I Murphy JD, Henderson K, Bowden 1I, Lewis M, Cooper GMI. Btipivacaine versus bupisacaine plus fentanyl for epidural analgesia: effect on maternal satisfaction. B.IM] 1991;302: 564-7. '9 March.:) 2 O'Driscoll K, Mleagher D. Active management oftlabour. London: Bailli&re Tindall, 1986. 3 National Maternity Hospital. Clinical report for the v'ear 1989. Dublin: National Maternity Hospital, 1989.

SIR,--It seems that the two main conclusions of Dr J D Murphy's study on epidural analgesia

during labour are that there was greater maternal satisfaction and fewer operative deliveries in women receiving epidural bupivacaine and fentanyl versus bupivacaine alone.' Both are in question as the inherent difficulties in this type of study were not overcome. The woman and the anaesthetist were aware of whether she was receiving one or two drugs. The use of an analgesic depends on the patient's and the anaesthetist's perception of the effectiveness of the drug. Thus, where the principal aim of a study is to assess a soft end point, such as maternal satisfaction, more rather than less effort must be made to blind the receiver and deliverer of the drug(s). Furthermore, no statement was made to suggest that the anaesthetists (preferably few) were standardised. It can only be assumed that the anaesthetists concerned were whomever happened to be "on that night." It is also very unlikely that the obstetricians who "were not told which treatment each mother had received" could not have spotted (and indeed needed to know) which drug was being given. The obstetrician's decision to proceed to operative delivery is based in part on maternal comfort. The important difference between the operative deliveries is that the two groups cannot therefore be attributed to the differences in analgesia, as was suggested. A G STEIN

South Croydon, Surrey CR2 6QF I Murphy JD, Henderson K, Bowden Ml, Lewis M, Cooper (iM. Bupivacaine vcrsus hupisacaine plus fentanyl for epidtural

analgesia: effcct oni matcrnal satisfaction. BMJ 1991;302: 564-7. (9 March.

AUTHORS' REPLY,-We are aware that the likelihood of a normal vaginal delivery is disappointingly low for primigravid patients who choose epidural analgesia in our hospital but without altering obstetric practices, which have an obvious influence, we have shown that the anaesthetists' management of epidural analgesia can also affect the mode of delivery. In answer to Dr Stein's letter, the mothers, obstetricians, and the anaesthetists who interviewed mothers after delivery were all unaware of the epidural treatment used. The epidurals were inserted and managed by the first three authors of our paper. Dr Stein's assertion that maternal comfort may have influenced the obstetricians' decisions about operative delivery is not supported as we showed similar analgesia in both groups of mothers. J D MURPHY K HENDERSON

M I BOWDEN M LEWIS

G M COOPER Department of Anaesthetics, Birmingham Maternitv Hospital, Queen Elizabeth Medical Cenitre, Birmingham B15 2TG

Organisation of antenatal care SIR,-In his discussion of the various styles of antenatal care' Professor Geoffrey Chamberlain did not mention community antenatal care schemes. Although perhaps not numerically important, they have brought benefits to both mother2 and baby' and have helped with administrative difficulties such as late booking4 and high defaulter rates.' In Huntingdon, where all the consultants visit general practitioners in the area, the perinatal mortality rate in 1987-9 was 4 6, the lowest in the country." The Tower Hamlets community antenatal care scheme with which we are involved has not only helped make antenatal care more convenient and personal for women but also produced obstetric

BMJ VOLUME 302

13 APRIL 1991

results comparable to and in some ways superior to more traditional forms of care (W D Savage and R M Cochrane, Nivel international conference on primary care obstetrics and perinatal health, s'Hertogenbosch, 1991). We accept that such a scheme, in which hospital specialists visit local surgeries during their antenatal clinics and in which women therefore need not attend hospital at all, would be inappropriate in some areas. Nevertheless, some may wish to emulate the idea for the benefits outlined above and for the fruitful interprofessional respect that such care engenders. WENDY SAVAGE RUTH COCHRANE SIVA PATHASUNDARAM Academic l)epartment of (iencral Practice and Primars Care, Medical Colleges of St Bartholomcw's and The London Hospitals, London EC 1M 6BO I Chamberlain G. ABC of antenatal care: orgamisationi of antenatal

care. B.lfj 1991;302:647-50. (16 Mtarch.) 2 ravlor R. Satellite clinics in maternity care. Alidwife. Health I isilor and Communitv Nurse 1986;22:287-8. 3 McKee 1. Community antenatal care: the Sighthill community atitenatal care scheme. In: Zander L, Chamberlain G, eds. I're:gnancy%' care for the 1980s. London: RSM, MacMillan, 1984: 19-38. 4 Robson J, Boomia K, Sasage W. Reducing delay in booking for antenatal care. ] R Coll Gen Pract 1986;36:274-5. 5 Newall Al, t)aw EG. Setting up tteighbourhood antenatal clinics: problems cncounitered in the inner city. Practitioner 1984;228: 1089-9 1. 6 Attonymous. Infant and perinatal mortality 1989: l)HAs. OPCS Monitor 1990;No 1, OH3 90/1.

Reporting of fine needle aspiration SIR,-We have three years' experience with fine needle aspiration-at a weekly surgical breast clinic. We would agree with Mr J Michael Dixon that the method is accurate, and prompt reporting allows repeat aspirations if the material initially obtained is inadequate. It reduces the patient's anxiety and avoids unnecessary return outpatient visits. There are many additional advantages not mentioned by Mr Dixon. Fine needle aspiration is a cost effective technique. It facilitates education of junior surgeons in aspiration technique as the cytopathologist provides immediate feedback as to the adequacy of material obtained and will, if required, demonstrate the correct technique of aspiration and preparation of material. A trained counsellor is in attendance and can provide immediate counselling to patients with suspicious or malignant lesions. The presence of a cytopathologist at the clinic provides a useful opportunity for discussion of difficult cases. The aspiration can be performed by either the surgeon or the cytopathologist and this allows a speedier throughput of patients. In general, however, the surgeon will perform the aspiration. We disagree with the statement that a "technician and an experienced cytopathologist have to be available to stain and report the findings." With the Diff Quik stain both staining and reporting are easily done by a cytopathologist. Mr Dixon wonders why the system is not more widely used in Britain. The technique is time consuming, requiring the cytopathologist to attend at the surgical clinic for several hours each week. There is at present a shortage of experienced cytopathologists, and heavy routine work commitments in the laboratory prevent many from being available to provide what is undoubtedly a valuable service. MARIA NAYAGAM JOAN M McCUTCHEON A E YOUNG St 'IThomas's Hospital, London SE 1 7EH

JMNI. Immediate reporting of fine needle aspiration of breast lesions. BM_7 1991;302:428-9. (23 February.)

I Dixon

BMJ

VOLUME

302

13

APRIL

1991

Diagnosing breast carcinoma in young women

cvtological diagnosis of breast cancer in young women. Br] Surg 1989;76:835-7. 4 Barrows GM, Anderson TJ, Lamb J, Dixon JM. Fine needle aspiration of breast cancer: relationship of clinical factors to cytology results in 689 primary malignancies. Cancer 1986;58:

SIR,-Based on a 19 year review of the diagnosis of breast carcinoma in young women, Dr A Yelland and colleagues seem to recommend that all discrete lumps in young women should be excised.' Diagnostic techniques have clearly improved over recent years and conclusions from this review must be questioned. The authors have confirmed our previously published findings that clinical examination and mammography have a low sensitivity for detecting breast cancer in young women.' Although they indicate that ultrasonography has been used in their centre, they do not present any results for this investigation, which is disappointing as this investigation is clearly useful in young women, in whom the sensitivity of mammography is low.' Of greatest concern, however, are their results for fine needle aspiration cytology. We have looked at the factors influencing the accuracy of fine needle aspiration cytology and have not found age to be an important factor.24 It is clear from our results that the accuracy of fine needle aspiration cytology has increased significantly over the past decade. Of the last 30 women with palpable breast cancer under the age of 36 years treated in our unit over the past three and a half years, all had fine needle aspiration and the findings reported as malignant in 27, suspicious in two, and acellular in one. If the unsatisfactory specimen is included then this gives a sensitivity for fine needle aspiration cytology of 97%. Sensitivity is usually calculated after excluding unsatisfactory aspirates, and if this is done the sensitivity for breast cancer in young women treated in our unit is 100% (compared with 78% reported by the group at St George's). These results are similar to those published recently by the group from Southampton.' It would be important to know what the results of cytology were over the past five years in the unit at St George's. If they are similar to our own and those of the Southampton group then the whole message of the paper by Dr Yelland and colleagues is undermined. Our current management in young women with clinically benign breast lumps is to perform a fine needle aspirate on all women. If the fine needle aspirate confirms that the lesion is benign then the patient is reviewed. Most patients who have discrete lumpy areas are usually then reassured and discharged, and an unnecessary biopsy is avoided. This approach has been shown to be safe.' Patients who have a clinical and cytological diagnosis of a fibroadenoma are offered the opportunity of having their lump removed under local anaesthesia or of keeping the lump under observation. Those patients who elect to keep their lump then undergo an ultrasound examination, which allows confirmation of the benign nature of the lesion and measurement of its size. The lesions are then scanned at three months, six months, and one year as we are currently monitoring the course of fibroadenomas. Increase in size is an indication for excision, but as yet we have not identified any patient whose lump has got bigger and therefore had to be excised. Over 90% of patients are currently opting for observation only, and this figure is similar to that reported in a study from Oxford.' The view that young women with breast lumps wish to have them excised is therefore clearly incorrect.

1493-8. 5 Dixon JM, Clarke PJ, Crucioli V, Dehn 'I'CB, Lee ECG, Greenall MJ. Reduction in the biopsy rate in benign breast disease using fine needle aspiration cytology with immediate reporting. Br] Surg 1987;74:1014-6.

J MICHAEL DIXON Royal Infirmary, Edinburgh EH3 9YW' A, G.raham MsD, Trott PA, c1t al. Diagnosing breast carcinoma in young women. BMJf 1991;302:618-20. (16 March.) 2 Dixon JMW, Anderson TJ, Lee D. Stewart HJ. Fine needle aspiration cytology in relationship to clinical examination and mammography in the diagnosis of a solid breast mass. BrJ Surg 1954;74: 593-6. 3 Ashley S, Royle GT, Corder A, et al. Clinical, radiolttgical and

1

Ytelland

AUTHORS' REPLY,-We compliment Mr J Michael Dixon and his colleagues on the extreme accuracy of diagnosis of cancer of the breast in patients under 36 in a small series of 30 patients. However, we would respectfully draw his attention again to the last paragraph of our paper, which states: "We suggest that centres not possessing adequate cytological and combined mammographic facilities should excise all discrete breast masses in this age group without previous investigation. We also suggest that the poor detection rate by general practitioners warrants all young patients presenting with a breast lump being referred to a surgeon with an interest in breast disease."' We think that it is a matter for discussion as to whether one can safely leave a presumed fibroadenoma in the breast with the added anxiety of repeated assessment and evaluation. We agree that most patients who have discrete lumpy areas can usually be reassured after evaluation. However, we were discussing a discrete mass. The workload of long term follow up in our unit would become prohibitive. Finally, a 100% accuracy in diagnosing carcinoma by cytological, radiological, and clinical means is a laudable aim but, in our opinion, is rarely achieved. R C COOMBES H T FORD J-C GAZET St George's Hospital, London SW 17 OQT 1 Yelland A, Graham MD, Trott PA. Diagnosing breast carcinoma in young women. BMJ 1991;302:618-20. (16 March.)

Compensation for medical accidents SIR,-The article by Mr Brian Capstick and colleagues on compensation for medical accidents raises a number of interesting issues-dealth with, adequately I believe, by my NHS (Compensation) Bill. He raises the problem of causation. It has been asserted by those opposed to the introduction of a no fault scheme that victims of medical accidents have as much difficulty in establishing causation as they do in proving negligence. As a result, they argue, no fault compensation will not assist them as they will still be required to prove a causal link between medical care received and resulting injury. I, and all those who supported my bill, dispute this. Unless one introduces a general disability compensation scheme, compensating individuals on the basis of need and not on the basis of how they acquired their disability, there will always be a need to prove causation. There will always be some worthy cases that do not qualify. The main hurdle under the present tort based system is the need to prove negligence. Many injuries result from "reasonable" care. They deserve compensation. My bill would have ensured that. My bill would not have resulted in a flood of trivial claims or a massive increase in costs. The 1948 Law Reform (Personal Injuries) Act would have been amended to ensure that when awarding damages a court could have taken into account the availability of NHS care rather than having to base an award on provision of private care. The bill also provided for periodic payments of awards. The Medical Injury Compensation Board established

911