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: Effect of flutamide on survival in patients with pancreatic cancer: results of a prospective, randomised, double blind, placebo controlled trial Brian A Greenway. 316:doi 10.1136/bmj.316.7149.1935

It is reassuring to note that Mr Greenway reports the treatment of a series of patients with pancreatic cancer, where histological diagnosis was achieved in only 28.8%1. Recently, we have retrospectively reviewed 13 patients with localised pancreatic cancer treated here with chemoradiation between February 1994 and July 1997. In our series histological diagnosis was achieved in 46.1%. All of our patients received 45Gy external beam radiotherapy in 25 fractions over 33 days together with 5FU 350mg/m2 and folinic acid 20mg/m2, 1 hour prior to the radiotherapy during the first and last five days of treatment. Maximal toxicity was noted as WHO grade 2 nausea / vomiting in 15%. 5 (45.4%) had image evidence of response. The median symptom free interval was 12 months. Excluding two patients who died within 6 weeks of completing treatment, 45.4 % (5/11) are still alive, 3 having no evidence of disease recurrence to date, 18, 18 and 21 months after diagnosis.

Histological diagnosis is normally considered imperative prior to radiation or cytotoxic treatment but given the above results, we have felt uncomfortable delaying effective chemoradiation for a positive histological diagnosis and like Greenway, rely on a constellation of signs and symptoms. The pancreas is difficult to image by conventional means, the gold standard still being spiral CT with intravenous and oral contrast2; small tumours are often more apparent at ERCP, especially those located in the head of the pancreas. 11 of our patients had assessable disease on CT, the remainder were confirmed visually at ERCP.

Endoscopic, ultrasound-guided fine needle aspirates have been reported to yield histological proof of malignancy in 53% 3 although we understand that in practice, this procedure can make stent insertions, with consequent symptom relief, impossible. In addition, a negative result cannot exclude malignancy and the smaller, potentially more curable lesions are those most likely to be missed. However, cytological examination of pancreatic duct brushings or catheter aspiration (during ERCP) can improve the rates of positive diagnosis, up to 90% if p53 immunocytochemistry is employed 4. The tumour associated antigen, CA19-9 can also be a useful adjunct to diagnosis, being elevated in 70-80% of patients with adenocarcinoma of the pancreas5.

We currently believe that we should continue to offer chemoradiation, if necessary on the basis of a constellation of signs and symptoms, but would like to urge all endoscopists to at least attempt pancreatic duct brushings and CA19-9 estimation prior to oncology referral. We also look forward to incorporating flutamide into our management of pancreatic cancer, preferably in the setting of an appropriate clinical trial.

Yours sincerely,

Dr. Marcia Hall, Senior Registrar, Medical Oncology
Dr. Rob Glynne Jones, Consultant in Clinical Oncology
Mount Vernon Hospital Cancer Centre

1. Greenway BA. Effect of flutamide on survival in patients with pancreatic cancer: results of a prospective randomised, double blind, placebo controlled trial. BMJ
1998;316:1935-8 (27 June)
2. Wyatt SH, Fishman EK. Spiral CT of the pancreas. Semin Ultrasound CT MR 1994;15:122-132.
3. Binmoeller KF, Thus R, Rathod V, Hanke P, Brand B, Jabusen HC, Soehendra N. Endoscopic ultrasound-guided, 18-gauge, fine needle aspiration biopsy of the pancreas using a 2.8mm channel convex array echoendoscope. Gastroinestinal Endoscopy 1998;47(2):121-7
4. Ishimaru S, Itoh M, Hanada K, Tsuchida A, Iwao T, Kajiyama G. Immunocytochemical detection of p53 protein from pancreatic duct brushings in patients with pancreatic carcinoma. Cancer 1996;77(11) :2233-9.

5. Steinberg WM, The clinical utility of the CA19-9 tumour associated antigen. Am J Gastroenterol 1990; 85:350-5.

Competing interests: No competing interests

11 August 1998
Marcia Hall
Senior Registrar
Re: Cochrane Injuries Group Albumin ReviewersWhy albumin may not work Abi Berger. 317:doi 10.1136/bmj.317.7153.235

The recent publications on this topic (1, 2) contain much of interest for us all and should give as all pause for though about the grounds both on which clinical practices in resuscitation are based and also on the methods by which success or failure in this area can be assessed. In addition there arises the question the validity of the method of systematic review of trials of varying quality - a topic referred to already in the considerable correspondence arising after the initial publication (1). Perhaps the complexity of the situation can be best considered by a more detailed look at the trials reported in the Cochrane Injuries Review (2) on burn injuries. There is little doubt that there is no consensus on the details of the best method for the resuscitation of burned patients, other than that is
an essential part of the initial management regime. That this is so is well illustrated in the latest reference work on burns (3). It is evident from the outset that the problem is not simple in view of the extreme variability in presentation and complexity of the injury, the progressive nature of the injury and the extreme susceptibility for
infection and metabolic disturbances in the more extensively burned patients. The authors of the report have considered the two most convincing trials on burned patients - one on adults (4) and one on children (5). The third trial considered, consisting of 14 patients in all (6), would probably not be accepted as convincing by a majority of
burn clinicians, in view of its distance from standard practice in burn resuscitation. The adult trial (4) shows clearly that colloid treated patients had greater lung water retention than did crystalloid patients at post burn day 7. Whether this is a contributory factor in the death of patients is not recorded, nor is the time of death or the cause of death in patients who died, in either group. The second trial (5) involving burned children, is in actuality a 'low albumin' versus 'high albumin' trial, the aim of the trial being to assess the value and efficacy of albumin supplementation to boost serum albumin levels in the stages after initial resuscitation. This trial is not actually
aimed at assessing the value of albumin as an agent for blood volume expansion in resuscitation and albumin supplements were given to the low albumin group to maintain serum albumin levels above 15 gm/L. An analysis of causes of death is not given and the discussion for the paper specifically refers to complications due to inhalation injury which may have effected death rates in the two groups. In the study the authors state that the trial examines the question of whether albumin should be supplemented and they are not conducting an evalaution of the value of albumin in resuscitation. The authors of these two reports(1,2) are to be complimented upon the wide range of their survey and the systematic and critical manner in
which they have conducted it. However meta-analyses will, in general, run the risk of drawing apparently general conclusions which may not be applicable to all the areas in the non homogeneous base from which the data is collected.. How this can be avoided is difficult to see and such problems may diminish the value of the analysis as a practical exercise. Perhaps further studies in this area should focus upon success in the achievement of clinical objectives (if these can be defined) rather than the crude statistic of mortality, unless a strong causative rather than associative connection can be established between
albumin usage and mortality. It could be argued that later mortality, especially in burn injured patients, is due to more powerful factors such as infection than the exact nature of the resuscitation technique and fluid used. We should also remember that some colloid-containing fluids, in particular fresh frozen plasma for children, may be helpful
in other ways than simply by acting to maintain and expand blood volume. We should not rush into precipitate abandonment of a useful material unless we have very good reason to do so. Whether a systematic review of this nature on such a wide ranging scale provides good enough reason to abandon albumin universally will have an interest beyond that of the immediate result of the analysis. Or will economics simply dictate the final result?
Let us see more systematic reviews of this type. It can never be wrong to challenge accepted practice. Either way the cookie crumbles, the result can only be beneficial but we must be sure that the underlying basis of the review is strong enough for everyone in the area of clinical practice from which the review is drawn to have confidence in
its findings.

Yours sincerely

Dr P G Shakespeare, PhD., Clinical Scientist,
Laing Laboratory, Salisbury District Hospital, Salisbury SP2 8BJ


1) Schierhout, G & Roberts I. Fluid resuscitation with colloid or crystalloid solutions in critically ill patients: a systematic review of randomised trials. BMJ, 1998; 316, 961-966
2) Cochrane Injuries Group Albumin Reviewers, BMJ, 1998; 317, 235-240
3) Warden GD. in Total Burn Care (Herndon DN, et al) W.B. Saunders Co. Ltd. ISBN 7-7020 -1827-9. 1996 Cap 6, 53ff
4) Goodwin CW, Dorethy J, Lam V, Pruitt Jr BA. Randomized trial of efficacy of crystalloid and colloid resuscitation on hemodynamic response and lung water following thermal injury Ann Surg 1983; 197, 520-531
5) Greenhalgh DG, Housinger TA, Kagan RJ et al. Maintenance of serum albumin levels in paediatric burns patients: a prospective, randomized trial (with discussion). J Trauma 1995; 39, 67-73 (Discussion 1995; 39, 73-4.)
6) Jelenko C (3rd), Williams JB, Wheeler ML, et al. Studies in shock and resuscitation. I: Use of a hypertonic, albumin-containing, fluid demand regimen (HALFD) in resuscitation. Crit Care Med 1979; 7, 157-167

Competing interests: No competing interests

Re: Perinatal death associated with planned home birth in Australia: population based study Hilda Bastian, Marc J N C Keirse, Paul A L Lancaster. 317:doi 10.1136/bmj.317.7155.384

> Re Perinatal death associated with planned home birth in Australia: population based study
> Hilda Bastian, Marc J N C Keirse, Paul A L Lancaster, BMJ 1998;317:384-388
> Bastian et al have shown the importance of audit of in maternity care to highlight adverse outcomes. The quality and relevance of antenatal care is important in the outcome of pregnancy and the choice of place of delivery is
> likely to dictate the type of antenatal care given. Good antenatal care should anticipate problems and activate timely transfer. Almost half of the stillbirths in their population based study were already dead before transfer
> had been effected, which shows that the transfer was too late. This alone emphasises the importance of timely referral. With the concept of proportional audit of perinatal mortality, which I introduced in 1996 responsibility
> for outcome is appropriately shared between the groups of carer when transfer takes place. When transfer is timely it seems unfair to attribute an unfavourable outcome to only one care facility. This encourages the timely
> transfer of care when problems are anticipated and provides a fair and helpful analysi!
> s of the morbidity and mortality which is associated with each form of care.
> David J R Hutchon
> Consultant Obstetrician, Memorial Hospital, Darlington.
> Hutchon D J R A method of proportional audit of perinatal care. Br J O G 1996:103;402-4

Competing interests: No competing interests

10 August 1998
D J R Hutchon
Consultant Obstetrician & Gynaecologist
Memorial Hospital, Darlington.
Re: Someone else's problem . 317:doi 10.1136/bmj.317.7152.194

Your anonymous correspondent ("Someone else's problem", filler, 18 July, p194) was upset by the attitude of some hospital staff when they seemed to regard Annie - a doubly incontinent patient from a neighbouring psychogeriatric hospital (who finally had a four hour operation on her "huge infected pressure sore") - as little more than a burden, obstructing and delaying the needs of others. But should we not perhaps be equally upset by the sickening description given of the forcible dressings that this poor demented patient repeatedly tried to refuse?

When, if ever, is it right to do others what we would not want done to ourselves? How many of us would sign an advance directive that "in the event of my old age being accompanied by dementia and double incontinence and a large infected bed sore, I wish daily dressings to be done by force if necessary, however much I resist, and I want this to be continued for as many weeks or months as it takes to prepare me for skin grafting"?

Competing interests: No competing interests

10 August 1998
Thurstan Brewin
Former Clinical Oncologist
Re: Internal inquiry clears surgeon Clare Dyer. 317:doi 10.1136/bmj.317.7155.367a

Dear Sir,

As reported in the BMJ (Aug 8), a local enquiry has now cleared Derriford Hospital's thoracic surgeon of any misconduct. During a recent visit to the Hospital I had the distinct pleasure of meeting Mr Rahamim and heard first hand of the effects of allegations made against him.

The copies of newspaper articles, internal memoranda and numerous testimonials pinned to noticeboards around the hospital leave the visitor in no doubt that Mr Rahamim has the admiration and support of the Hospital Board, his colleagues and, most importantly, his patients. There appear to be no doubts about his clinical ability locally.

Assuming that an independent enquiry reaches the same conclusion as that conducted locally, what will be the long term consequences? In the wake of the Bristol case and after the uncomfortable experience of being senselessly placed in the media spotlight, how can any surgeon practice with the same degree of confidence? For how long will a surgeon in this situation be reluctant to take on difficult or 'high-risk' cases for fear of further criticism?

Ultimately the quality of patient care may suffer. Is the style of investigative journalism exhibited in this case really in the public interest? I think not!

Competing interests: No competing interests

10 August 1998
Joseph E Arrowsmith
Visiting Associate in Anesthesiology
Duke University Medical Center, Durham, NC 27710, USA
Re: Cochrane Injuries Group Albumin ReviewersWhy albumin may not work Abi Berger. 317:doi 10.1136/bmj.317.7153.235

EDITOR - Roberts et al have written what is termed a ‘systematic review’ of albumin administration in the critically ill, with the sole inclusion criterion being that trial subjects had to be prospectively randomised ‘to one of two (or more) interventions’.1 But amongst a group of 20 trials where albumin was used to treat hypovolaemia, there are five trials (totalling 26% of patients) where the controls were simply given ‘no albumin’ implying no intervention at all. In these studies presumably either the control group were left grossly under-resuscitated or the intervention group suffered iatrogenic fluid overload; either scenario raises questions about the ethical conduct of these studies and casts doubt on any conclusions drawn in relation to mortality.

The above point is made to illustrate the problems inherent in omitting any critical analysis of the original studies. Unfortunately, the contributor list headed by Roberts appears to lack individuals capable of such analysis - particularly glaring is the absence of anyone with adult intensive care expertise. This lack of clinical insight was also identified as a problem by respondents to a forerunning article from the same group, a meta-analysis of fluid resuscitation with colloid or crystalloid.2

However, on this occasion the authors have done more than merely publish a paper of dubious scientific merit - they have also actively subverted the peer review process. Firstly, Roberts et al are reported to have urged the Health Secretary to take ‘appropriate steps to protect the public’3 six weeks before the BMJ had even published their article. They then went on to participate in a classic media circus which will have created needless anxiety for many patients and needless confusion for many clinicians. The net result is that Roberts et al have made a mockery of their own call for albumin use to be confined to the context of a ‘rigorously conducted randomised controlled trial’.1 Such a trial will of course now be virtually impossible. The process involved here has had nothing to do with evidence based medicine and it is unfortunate that the potential of the Cochrane Collaboration is being misused in this way.

Mark R Nel Senior registrar in anaesthetics
Hammersmith Hospital, London W12 0HS

1 Roberts I et al. Human albumin administration in critically ill patients: systematic review of randomised
controlled trials. BMJ 1998;317:235-40. (25 July.)
2 Schierhout G, Roberts I. Fluid resuscitation with col-loid or crystalloid solutions in critically ill patients: a systematic review of randomised trials. BMJ 1998;316:961-4.
3 300 die as health chiefs dither. The Observer. London: 26 July 1998.

Competing interests: No competing interests

10 August 1998
Mark R Nel
Senior registrar in anaesthetics
Hammersmith Hospital, London W12 0HS
Re: Nuchal test detects 80% of Down's syndrome pregnancies Jacqui Wise. 317:doi 10.1136/bmj.317.7155.368b

Nuchal test detects 80% of Down's syndrome pregnancies

The Snijders study referred to in this commentary reports an 80% detection rate, but using the same method of analysis others have not been able to emulate as high a rate. The explanation may lie in the expertise of the ultrasound or the organisation of the service. Another explanation is that we may be dealing with different populations. Is nuchal thickness related to fetal size or to fetal maturity? Mittendorf et al have demonstrated useful differences in the normal length of pregnancy according to maternal age, parity and racial origin. Current ultrasound dating relates to the start of the pregnancy but places no attention on the precise length of the pregnancy. Hermanni Boerhaave related interval to delivery for his "dating marker" - the LMP. I propose a similar method using ultrasound. The methodology allows the generation of fully customised charts for maternal age, etc. If nuchal thickening is related more to maturity than to fetal size, the sensitivity of the test may well be improved in different sub-populations.

The use of electronic publishing allows the ready publication of raw data along with the study details. This should be encouraged as it permits readers to confirm the researchers' analysis of the data.
David J R Hutchon
Consultant Obstetrician,
Memorial Hospital, Darlington.

Haddow JE Antenatal screening for Down's syndrome: where are we and where next? Lancet 1998;352:343-6.

Hutchon DJR "Back to the Future" for Hermanni Boerhaave or a rational way to generate ultrasound scan charts for estimating the date of delivery.

Competing interests: No competing interests

10 August 1998
D J R Hutchon
Consultant Obstetrician
Memorial Hospital, Darlington
Re: The 1998 European Resuscitation Council guidelines for adult single rescuer basic life support . 316:doi 10.1136/bmj.316.7148.1870

EDITOR - Instruction in cardiopulmonary resuscitation (CPR) has become a standard part of training for medical personnel and is widely recommended for the lay public. However, concerns regarding mouth-to-mouth ventilation appears to create substantial barriers to performance of CPR in and out of hospital setting. This is especially true if there is visible blood around the mouth of the victim.

Transmission of HIV during mouth-to-mouth ventilation requires that three conditions be fulfilled: (1) an HIV positive victim, (2) blood-blood contact, and (3) seroconversion of the rescuer. Knowing the prevalence of each one of these independent risk factors, it is possible to calculate the chance of contamination. The HIV prevalence in United Kingdom is about 0,0004%. Considering that the victim has visible blood around the mouth, blood-blood contact will be possible if the rescuer possesses a microlesion in his/her oral mucosa. Such microlesion may be present in up to 50% of healthy persons . Seroconversion happens in 0,3% of the cases of blood-blood contact. Therefore, the chance of contamination is approximately one in a hundred milion (0,0004x50x0,3). This is a hundred thousand times lower than the risk of anaphylactic shock after the use of penicilins. Also, not a single case of transmission of HIV by mouth-to-mouth ventilation was reported, while there are two cases of persons probably infected by kiss route.

The very low frequency with wich HIV has been transmitted trough oral-to-oral route is impressive. This is probably due to the risk of infection may depend on the size of the innoculum, wich is frequently small by oral-to-oral route. The finding of inhibitory substances in saliva may further reduce the probability of infection and cannot be considered in our numerical estimatives.

Educating about the percent of patients that survive cardiac arrest, the actual risks of contracting HIV, and preventative measures, should result in an increased willingness of laypersons and physicians to perform mouth-to-mouth ventilation on victims of cardiac arrest.

Competing interests: No competing interests

10 August 1998
Carlos Frederico Arend
Medical student
Federal University of Rio Grande do Sul - Brazil
Re: When can odds ratios mislead? Huw Talfryn Oakley Davies, Iain Kinloch Crombie, Manouche Tavakoli. 316:doi 10.1136/bmj.316.7136.989

Dear Sir,

In a recent article, Davies et al. (1) commented on a potential problem when interpreting odds ratios (OR) as relative risks (RR) in epidemiological studies. However, their vague concept of effect measures as applied to different study designs in epidemiology may lead to misuse and false interpretation of OR.

Davies et al. (1) state that the odds ratio is a common measure in case-control studies, cohort studies, or clinical trails. Unfortunately, this first sentence of their article is not correct. For different study designs, OR should only be used as a measure of effect size when RR can not be estimated directly. In cohort studies as well as in clinical trials, RR (the cumulative incidence ratio or the incidence density ratio) can be estimated directly. Therefore, there is no need to use OR to represent the effect size. In contrast in case-control studies, incidence data is usually not available. Therefore, the ratio of the odds of exposure among cases to the odds of exposure among non-cases is calculated. In theory, all case-control studies can be viewed as nested case-control studies, in which both cases and controls are drawn from a well defined source population. If the controls are selected by incidence density sampling, then the OR derived from the case-control study is, apart from random error, the same as RR in the source population. No rare disease assumption is needed (2). In fact, only when cumulative sampling in case-control studies is used, the rare disease assumption is needed. Usually this assumption should not represent a problem, because case-control designs typically are preferred when the outcome of interest is rare (say less than 5%).

The situation, however, is different in cross-sectional studies (usually applied to investigate more common outcomes) when the prevalence odds ratio (POR) is used as an estimate of the prevalence ratio (PR). Since in cross-sectional studies only prevalent cases are drawn, there is no direct way of estimating RR. In the general population, the prevalence odds is equal to the product of the incidence times disease duration (3). If we assume that the exposure of interest has no influence on the disease duration, then the POR is, theoretically, equal to the RR. However, the assumption of equal duration of disease among the exposed and unexposed population is often questionable. Therefore, some authors propose the use of PR as a conservative estimator of the RR (4).

1 Davies HTO, Crombie IK, Tavakoli M. When can odds ratio mislead? BMJ 1998;316:989-91.
2 Greenland S, Thomas DC. On the need for the rare disease assumption in case-control studies. Am J Epidemiol 1982;116:547-53.
3 Rothman KJ, Greenland S. Modern Epidemiology. 2nd ed. Philadelphia: Lippincott-Raven, 1998.
4 Thompson ML, Myers JE, Kriebel D. Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done? Occup Environ Med 1998;55:272-77.

Competing interests: No competing interests

10 August 1998
Dirk Taeger
Statistician (Dirk Taeger), Epidemiologist (Yi Sun, Kurt Straif)
Yi Sun, Kurt Straif
Institute of Epidemiology and Social Medicine, University of Muenster, 48129 Muenster, Germany
Re: Videos, photographs, and patient consent Catherine A Hood, Tony Hope, Phillip Dove. 316:doi 10.1136/bmj.316.7136.1009

Hope, Hood and Dove (1) rightly stress that "the internet and electronic
publishing are powerful tools for the dissemination of medical information
and have created a demand for medical images" and emphasise that patient
images should, in most circumstances, not be used without consent. However,
in the age of the Internet, the links between images, patients and medical
students and staff are complex and non-intuitive, so that the enormous
potential of the Internet and World Wide Web in medical education need
not be neglected because of ethical issues. Beginning with the issue of
security, Hope, Hood and Dove (1) state that "with current levels of security
on the internet, there is little to ensure that such images are not widely
seen, distributed, or misused." On the contrary, it is not difficult to
limit who can see what images over the web, either by limiting web-access
to a given image or page to certain IP addresses, subnets, or domains,
by using password protection or for very sensitive material, by using encryption
through public key cryptography (see The World Wide Web Security FAQ:
Of course, once medical students have downloaded digital images they could
pass them on to those outside the medical profession, however, to do so
would be to flout local and professional ethics in the same way as showing
confidential photographic images to their non-medical associates. With
the easy availability of image manipulation software, it is relatively
easy to create images that do not reflect a true likeness of any real individual.
For example, cover-girl images are commonly touched up to remove blemishes
or make the whites of their eyes whiter; the magazine Wired has featured
a cover image of O. J. Simpson as a near-Scandanavian caucasian (;
the Benetton magazine "Colors" once commissioned a picture of Ronald Reagan
suffering from AIDS, complete with multiple Kaposi's lesions (
and the Face Factory software allows one to assembly line drawings of faces
from parts of other people's faces (
We wished to see whether we could, in a single afternoon, create fictional
images of photographic quality illustrating medical conditions. We began
by attempting to create a malar butterfly rash such as one might see in
systemic lupus erythematosis on a face that does not exist in the real
world. To create the face, we used the morphing program MorphMan 95 on
a PC to combine the faces of the two authors. Next we used Photoshop 3.0
on a Macintosh to create the rash (Figure 1), repeatedly employing the
magic wand tool and adjusting the hue of selected areas.

Figure 1

When we showed this image to a medical colleague asking for the diagnosis,
he neither suspected that the rash was fake, nor did he recognise us in
the image. Next we attempted to create a fictional image of a melanoma.
We morphed together two melanoma images that we obtained from the Internet,
then placed the composite image onto an image of the arm of one of the
authors (MJP). We then used Photoshop's clone tool to alter the shape and
textural surface of the lesion (Figure 2).

Figure 2

In doing this, one could argue that we have used images without consent;
however as with combining research data, in combining images in this way,
the final image is probably so far removed from any one patient as to obviate
the need for consent. Had time permitted, we could have combined dozens
rather than just three images. In such cases how much of an image would
warrant consent--following a reductio ad absurdam, would use of a single
pixel require informed consent? We conclude that in the age of Lara Croft
and Kyoko Date (,
it is time for medical illustrators and educationalists to free themselves
from the constraints of real world images--the future is virtual!

Competing interests: No competing interests

07 August 1998
Mark Pallen
senior lecturer (MP), web resources devt officer (NL)
Nick Loman
St Bartholomew's and the Royal London School of Medicine and Dentistry