Search all rapid responses

All rapid responses

Rapid responses are electronic letters to the editor. They enable our users to debate issues raised in articles published on Although a selection of rapid responses will be included online and in print as readers' letters, their first appearance online means that they are published articles. If you need the url (web address) of an individual response, perhaps for citation purposes, simply click on the response headline and copy the url from the browser window. Letters are indexed in PubMed.

Re: Medical and psychosocial effects of early discharge after surgery for breast cancer: randomised trial Paul I M Schmitz, Marinus A Paul, Theo Wiggers, et al. 316:doi 10.1136/bmj.316.7140.1267


While one welcomes randomised trials, the study by Bonnema et al fails to answer the questions they have sought to address. Firstly, the two groups being compared contain a mixed cohort of patients undergoing breast-conserving surgery and modified radical mastectomy. They therefore have made the fundamental assumption that women undergoing both procedures behave exactly the same way in terms of their post-operative complication rate for eg, volume of seroma fluid drainage and suffer the same degree of psychological morbidity. They should have been stricter in their inclusion criteria and only recruited patients who underwent the same surgical procedure. This would have made their results more meaningful.

The concept of keeping patients in hospital for 9-12 days post-operatively is archaic. In our practice, which probably reflects the practice in the rest of the UK, the mean post-operative hospital stay is in the region of 4 days with drains being removed on day 5, irrespective of volume of drainage of fluid.

One of their aims was to address the complication rate following early discharge. However,in discussion, they state that "the number of patients in this study was too small to detect a difference of 5% in rates of wound complication" and subsequently claim that recruitment of 800 patients, which is what would have been required, would not have been "feasible in this type of research". Why?

It is vital that any study examining shorter hospital stay must involve a detailed analysis of costs with the help of a health economist to calculate in-hospital and community costs. This is particularly important for the UK where NHS funding is central.

The follow-up period to assess psychological morbidity is too short. At 3 months patients may be undergoing adjuvant therapy, loco-regional radiotherapy and systemic chemotherapy, which add to their morbidity. It is essential that such studies are designed to assess psychological morbidity at completion of treatment to provide a more meaningful result. In this study, a further set of questionnaires to be completed at one year would have been necessary.

These issues are currently being addressed in a randomised trail in our institution, funded by the Scottish Office, which will complete recruitment at the end of 1998. Results from this study will hopefully clarify all the issues raised above.

Competing interests: No competing interests

18 May 1998
A D Purushotham
Senior Lecturer in Surgery
Western Infirmary, Glasgow, G11 6NT
Click to like:
Re: Can students learn comparable clinical skills in general practice and hospital settings? . 316:doi 10.1136/bmj.316.7143.1531a

In response to the above article it is interesting to note that the Methods section does not state that ethical approval was sought from an independent body, nor does it tell us whether informed consent was obtained from all individual students. This raises issue of violation of student Autonomy by those in the medical school who arranged this trial while being responsible for their welfare.

yours sincerly,

Matthew Miller FRCS

Competing interests: No competing interests

18 May 1998
Matthew Miller
Senio SHO General Surgery
Victoria Infirmary, Glasgow
Click to like:
Re: Vulval Pain Society provides information on vulval symptoms David Nunns, Diane Hamdy. 316:doi 10.1136/bmj.316.7132.706b

Dear David Nunns

Your article on the Vulval Pain Society states that several support groups exist. Could you please advise me if a support group for suffers with vulval symptoms exists in Australia or particulary Western Australia.


Competing interests: No competing interests

18 May 1998
Sky Taluha
Click to like:
Re: New method of expressing survival in cancer is popular Jayant S Vaidya, Indraneel Mittra. 316:doi 10.1136/bmj.316.7137.1092

Nearly one year ago, Vaidya and Mittra proposed a new method that expresses survival in cancer patients by determining the "normal remaining life" of all individuals that belong to the cohort under examination (1,2). This method requires the calculation, for each subject of the cohort, of the ratio between survival after diagnosis and life expectancy after diagnosis (or, more in general, survival and life expectancy after the event that defines the time zero in the follow-up).

In comparison with traditional survival calculations, the change introduced by this new method intervenes at the level of handling individual patient data. In non-censored patients (e.g. patients who die), the new method constructs the foregoing ratio by placing measured survival at the numerator and remaining life-expectancy at the denominator. For example, if one assumes that diagnosis represents the time zero in the follow-up, the fractional survival time for each individual is calculated as: fractional survival time = (survival from diagnosis to death )/ (normal remaining life after diagnosis). This fractional survival time is then introduced in standard survival-curve calculations.

While there is no problem in managing non-censored patients through the above procedure, in the case of censored patients (e.g. patients who are alive at the date of closure of the study) the new method uses the same value at the denominator (i.e. normal remaining life after diagnosis), but is forced to use censored survival (i.e. survival from diagnosis to the last contact with the patient) at the numerator (because uncensored survival -i.e. survival from diagnosis to death- is unknown). Unfortunately, Vaidya and Mittra have not realised that censored survival can be a macroscopic underestimation of uncensored survival, particularly in young patients. In traditional survival-curve calculations, uncensored and censored survival times are not divided by any denominator and the subsequent statistics can appropriately handle terminations of follow-up. In the new method, survival times are divided by an appropriate denominator in cases of uncensored patients, but censored survival times are divided by a disproportionally large denominator; the subsequent statistics can try to handle terminations of follow-up, but the data already contain a heavy bias that tends to worsen the survival curve of cohorts containing many censored patients.

These arguments explain why Marubini and Mariani (3) have found a very small probability of cure (13%) in their cohort of 30-year-old, node-negative women. This 13% value, that was obtained by application of the new method, was probably an artifact that resulted from the use of censored survival instead of uncensored survival (that was unavailable). Fortunately, these young patients have a probability of cure much higher than 13% (even though no precise calculation can be made because of the censored nature of these data).

In conclusion, the new method proposed by Vaidya and Mittra (1,2) produces reliable results only when the investigators examine data where all subjects have died during the follow-up. Its application is grossly misleading when considering cohorts with censored patients.

The method of Vaidya and Mittra seems to be -in most cases- inapplicable, but the underlying idea is interesting. Further research is needed in devising other survival analyses that account for normal remaining life.

Andrea Messori, Sabrina Trippoli

Laboratorio di Farmacoeconomia
Drug Information Centre
Azienda Careggi
50134 Firenze, Italy (E-mail:


1. Vaidya JS, Mittra I. Fraction of normal remaining life span: a new method for expressing survival in cancer. BMJ 1997; 314: 1682-1684.

2. Vaidya JS, Mittra I. New method of expressing survival in cancer is popular BMJ 1998;316:1092

3. Marubini E, Mariani L. New method for expressing survival in cancer (Letter). BMJ 1997;315:1375-6.

Competing interests: No competing interests

18 May 1998
Andrea Messori
Coordinator, Drug Information Centre
Azienda Careggi Hospital, Florence, 50134 Italy
Click to like:
Re: Cost effectiveness of community leg ulcer clinics: randomised controlled trial Louise M L Brereton, Jean Peters, Charles G D Brooker, et al. 316:doi 10.1136/bmj.316.7143.1487

Compression bandaging is the most effective form of treatment for venous leg ulcer.The comparisons that really need to be made are between the different compression bandaging systems available,not as in this study between compression bandaging and a control group which is being treated with a range of what are clearly less effective interventions.

Competing interests: No competing interests

16 May 1998
John C Platt
nurse lecturer
Click to like:
Re: What makes a healthy website? Douglas Carnall. 316:doi 10.1136/bmj.316.7143.1542
In his review of health information on the BBC's website I note Douglas Carnell alludes to what he perceives as the "banality" of the site's women's health link. In my experience, a sizable proportion of the Web's consumer health sites could be described not only as banal but also as woefully impoverished of accurate material; and only a minority I believe exhibit an accordance with the Health On The Net Foundation Code of Conduct.

One of the principal factors contributing to this deficiency is that so many of these sites lack a broad input from formally qualified medical professionals, and UK doctors are particularly conspicuous by their absence. To me this is worrying.

Yes, we should accept that in this information age members of the public expect to have more profound knowledge of health matters - their avid use of the Internet to glean such information is surely testimony to that desire to know more. What we should not accept, as responsible medics, is that they earnestly seek answers only to receive erroneous or misleading advice.

The opportunities for UK doctors to personally contribute to public health information, however large or small these contributions, or to become involved in Health Informatics in general, are now widening. Sadly these are opportunities that are being missed, partly because of the existing inadequate support for doctors with a keen and growing interest in the speciality.

With the increasing emphasis placed on preventative measures in public health, Health Informatics, in its many forms, has a major role to play. Yet for UK doctors, especially junior doctors, wishing to progress in this field, either part-time or full-time, the path is a very difficult one. No formal career structure is offered, and, in spite of the developing awareness of the importance of Health Informatics by a number of medical bodies, there still remains a lack of readily accessible advice on such issues as research and relevant qualifications.

In response to the title of Dr Carnall's article, 'What makes a healthy website?', may I suggest that UK medical bodies consider with some urgency how they might improve the current situation for doctors involved in Health Informatics, thus encouraging greater interest from and involvement by the profession in general. Then in future years reviews of the Web's public health information might not be describe it as banal or inadequate.

Competing interests: No competing interests

16 May 1998
Jeremy Sims
Locum SHO Psychiatry
Farnham Road Hospital, Guildford; & Chairman, Cyber-Hospital
Click to like:
Re: Inappropriate use of sumatriptan: population based register and interview study Birthe K Rasmussen, Jakob Kragstrup, Lars F Gram, et al. 316:doi 10.1136/bmj.316.7141.1352
The problem of sumatriptan overuse has been raised by others(1, 2) and as reported by Gaist and colleagues(3) is of considerable concern, particularly in the context of Chronic Daily Headache(4). It has been suggested that sumatriptan may replace ergotamine in overuse syndromes(5) and indeed the issue of sumatriptan overuse has been raised in these pages(6). We have now seen in our clinic daily use for each of sumatriptan, naratriptan and zolmitriptan, and although thus far only sumatriptan overuse has been observed de novo without a switch from another acute anti-migraine compound, it seems reasonable to speculate that the problem may be seen with any triptan. The core of this problem, which deserves emphasis here, is that daily or near-daily use of any triptan is, with very few exceptions, inappropriate. Moreover, daily or near-daily use of any acute anti-migraine compound in use in the UK may lead to significant management problems, in particular analgesic-associated headache(7). The commonest cause of this we see in our Headache Clinic at the National is overuse of compound analgesics, particularly those which include codeine phosphate. If daily or very frequent headache is as common in the UK as it is in Spain(8) or the United States(9), at 4-5% of the population, overuse of various pain-killers is likely to be a considerable public health issue that has not hitherto received adequate attention. Our experience suggests that the issues raised by Gaist and colleagues(3) are the tip of the overuse iceberg that represents a modern and perhaps man-made epidemic.

1. Pini LA, Trenti T. Does chronic use of sumatriptan induce dependence? Headache 1994;34:600.

2. Catarci T, Lenzi GL, Cerbo R, Fieschi C. Sumatriptan and daily headache. J. Neurol. Neurosurg. Psychiatry 1995;58:508.

3. Gaist D, Tsiropoulos I, Sindrup SH, et al. Inappropriate use of sumatriptan: population based register and interview study. BMJ 1998;316:1352-1353.

4. Silberstein SD, Lipton RB, Sliwinski M. Classification of daily and near-daily headaches: a field study of revised IHS criteria. Neurology 1996;47:871-875.

5. Catarci T, Fiacco F, Argentino C, Sette G, Cerbo R. Ergotamine-induced headache can be sustained by sumatriptan daily intake. Cephalalgia 1994;14:374-375.

6. Kaube H, May A, Diener HC, Pfaffenrath V. Sumatriptan. BMJ 1994;308:1573-1574.

7. Mathew NT. Drug-induced headache. Neurologic Clinics 1990;8:903-912.

8. Pascual J, Castillo J, Guitera V, Munoz P. Epidemiology of chronic daily headache in the general population. Neurology 1998;50.

9. Scher A, Stewart WF, Liberman J, Lipton RB. Prevalence of frequent headache in a population sample. Headache 1997;37:330.

Competing interests: No competing interests

16 May 1998
Peter J Goadsby
Reader in Clinical Neurology/Wellcome Senior Research Fellow
Institute of Neurology, Queen Square London WC1N 3BG UK
Click to like:
Re: Identifying asthma and chronic obstructive pulmonary disease in patients with persistent cough presenting to general practitioners: descriptive study J C van Houwelingen, M P Springer, D S Postma, et al. 316:doi 10.1136/bmj.316.7140.1286
EDITOR -- Thiadens, et al examined 192 patients presenting to their general practitioner with persistent cough and discovered a high prevalence of asthma (39%) and chronic obstructive pulmonary disease (7%). These researchers and the editors of the BMJ are to be congratulated on tackling a common and challenging problem in primary care practice and applying it to scientific scrutiny. Articles such as this one are what maintain the BMJ at the forefront of medical literature and information.

I was disappointed that Thiadens et al did not provide a control group for examination, i.e. asymptomatic individuals in the community. Specifically, what is the prevalence of abnormal pulmonary function tests in their general population? If it is high, with a prevalence
approaching that found in this coughing population,
an alternative conclusion might be offered: that asthma and chronic obstructive pulmonary disease are common and not significantly more so in the population of coughers.

I encourage Thiadnes, et al to pursue further examination of their study group as I am sure that much more information can be garnered here. One issue deserving scrutiny is whether these patients are still coughing after six months and thereafter? Also, what diagnoses can be rendered (if any) to the remaining (54%) of coughers and does their prognosis differ from the group diagnosed with asthma and chronic obstructive pulmonary disease? Finally, how do these patients respond to different therapies?

Competing interests: No competing interests

16 May 1998
Joseph M Rothenberg
primary care internist
Jerusalem, Israel
Click to like:
Re: Can students learn comparable clinical skills in general practice and hospital settings? . 316:doi 10.1136/bmj.316.7143.1531a
It was kind of Elizabeth Murray and her colleagues to cite my book, Cross-over Trials in Clinical Research in their reply. Nevertheless, I have reservations about the use of such trials in assessing the effect of education. The essential feature of a cross-over trial is that subjects are allocated to sequences of treatments in order to study the effects of individual treatments(1). An necessary condition of using such trials is that the effects of treatments be reversible. This is quite plausible when comparing different beta-agonists in asthma, for example. In an educational setting, however, this would mean that students would have to forget what they had previously learned. I have enough confidence in the authors' educational abilities to doubt that this will have been the case.

1. Senn, S.J. Cross-over Trials in Clinical Research, Wiley, Chichester, 1997.

Competing interests: No competing interests

16 May 1998
Stephen Senn
Professor of Pharmaceutical and Health Statistics
University College London
Click to like:
Re: How should different life expectancies be valued? Norman Waugh, David Scott. 316:doi 10.1136/bmj.316.7140.1316

Diminishing marginal utility of health effects

EDITOR - Recently, Waugh and Scott wrote a letter in which they raised some important questions and proposed to triple or double health effects when total life expectancy is below 6, repectively 12 months taking the ‘duration of life time left’ into account in economic evaluations.1 We feel that they raised a number of issues, in which economic theory may be of help.

Firstly, the principle of attaching more weight to benefits gained when life expectancy is short, seems familiar. This idea corresponds to the economic principle of diminishing marginal utility, reflecting that giving an additional sandwich to someone having few is to be preferred over giving it to someone having many. Applying this principle here implies that giving an additional QALY to a person with a quality adjusted life-expectancy of 20 years, is less valuable than adding one to a person with an expectancy of only 3 months. That this notion implicitly is already used, may be derived from the fact that life-saving lung-transplantation, with a huge costs per QALY, is considered worthwile, while prevention programmes for cholestorol, with much lower costs per QALY are not considered cost-effective. Additionally, the same notion may explain the acceptance of high costs in last year(s) of life, where potential health gains and life-expectancy are oftentimes low. However, we do feel that Waugh and Scott’s proposal to triple or double health effects is as arbitrary as no adjustment and that more research is needed to find the appropriate weights.

Secondly, correcting for diminishing marginal utility may partly solve the problem that persons with a shorter life-expectancy may be more willing to accept a poor quality of life than persons with a longer life-expectancy. However, this also relates to one of the principles underlying QALYs, namely that of constant proportional trade-off, which means that indifference between 10 years in health state A and 5 years in health state B implies indifference between 10 months in A and 5 months in B. Again, more research is needed to indicate how the concept of QALYs should be adapted for situations involving very short life-expectancies.

Finally, in relation to the above, Waugh and Scott mention discounting of future effects. It should be noted that discounting and diminishing marginal utility are two different subjects, with similar consequences, but from completely different backgrounds. Therefore, we feel that they should be treated separately.

Werner Brouwer, Research fellow
Ben van Hout, Senior Research Fellow
institute for Medical Technology Assessment
Erasmus University Rotterdam
PO Box 1738, 3000 DR Rotterdam, The Netherlands
phone: + 31 10 408 8584, fax: + 31 10 452 6086

1 Waugh N, Scott D. How should different life-expectancies be valued? BMJ 1998: 316: 1316

(Word count: text including title 399 words)

Competing interests: No competing interests

14 May 1998
Werner Brouwer
Research Fellow (Brouwer) and Senior Research Fellow (Van Hout)
Ben van Hout
institute for Medical Technology Assessment, Erasmus University Rotterdam, The Netherlands
Click to like: