Bandolier Letters Page

Bandolier recognises that there are other points of view to those that appear in its pages, and from time to time we publish letters we receive with a different perspective.This page groups the letters Bandolier has received, with links to the original articles.
Alternative Medicine
Antimicrobials for Cystitis
Bandolier - a ladder for the lottery
CABG mortality
Cholesterol Screening
Fashion and evidence
Fungal Nail Infections
G.P's guide to home birth
Hospital-led Prescribing
Implementation gap
Hospital acquired infection Bando 73 February 2000
Interferon in MS
Nasal go-faster strips
Numbers-needed-to-treat (NNT) or Likelihood of Being Helped or Harmed (LBHH)?
Psoriasis & eczema review
Quality and size
Reading in bed
Screening blacklist
Shopfloor epidemiology
Shoulder Pain
Skeins and in-flight decisions
Teenage pregnancy
Third generation pills
Re: Bandolier number 55 - Evidence for Tonsillectomy
Tonsillectomy for sore throats: Vol 5, Issue 9, September 1998
Verrucas and Minor surgery - warts and all

Antimicrobials for Cystitis
Drs Martin and Pande and Mrs Ainsworth, from Fitznells Manor Surgery, Ewell, Surrey, wrote in response to Bandolier 13 . Our own mea culpa is in Bandolier 15 .


In UK general practice the choice of antibiotic is often made, and treatment commenced, before the laboratory report on the susceptibility of the causative organism is available. Knowing the probable infecting organism and its likely susceptibility is required. It is essential to base prescribing policy on local statistics in view of geographical variations which are known to occur.

A recent review of antibiotic policies and their relevance to general practice prescribing suggests that in the UK rational antimicrobial prescribing requires effective communication between the FHSA, microbiologists and GPs. Antimicrobial prescribing guidelines in the treatment of urinary tract infection were produced in Leicestershire and Derbyshire. The Leicestershire guidelines, based on local susceptibility data, drew attention to the substitution of trimethoprim alone for Co-trimoxazole.

In Epsom and Ewell in Surrey, 83% of urinary isolates from GPs (total 4082) in 1994 were coliforms. A breakdown of antibiotic sensitivities to coliforms showed 58% sensitivity to amoxycillin, 76% to trimethoprim, 90% to nitrofurantoin, 95% to cephalexin and 100% to ciprofloxacin (S Chambers, PHL Epsom, personal communication).

A cost analysis reveals that if the first-line treatment of patients with a coliform UTI (3394 isolates) was trimethoprim 200 mg twice a day for five days (cost of course £0.38), this would cost the locality £1,289. If the remaining 24% of patients who were resistant were then treated with cephalexin 500 mg three times daily for five days (cost of course £3.60) the combined cost would be £4,220. The cost of using cephalexin as first-line on all patients would be £12,218.

Both treatments would leave about 5% of patients resistant to treatment, though using trimethoprim first-line would result in saving £7,998 for the treatment of urinary tract infections in Epsom and Ewell. Given the development of locality purchasing, such cost analyses are becoming increasingly important.

In broad terms this suggests that to ensure an extra 20% of women receive an antibiotic to which they are sensitive would cost three times as much, against the price of a few extra days of discomfort.

We realise, of course, that sensitivities are in-vitro findings and may not always correlate with clinical response. However, the main message must be that rational prescribing of antimicrobials (that which is effective, appropriate, safe and economical) requires knowledge of local data, effective communication between GPs and microbiologists, and knowledge of the safety profiles of the drugs used.
Dr RM Martin GP, Dr S Pande GP Trainee, Mrs V Ainsworth In-house Pharmacist.
Fitznells Manor Surgery
Ewell, Surrey.


TM Hooton, C Winter, F Tiu, WE Stamm. Randomized comparative trial and cost analysis of 3-day antimicrobial regimens for treatment of acute cystitis in women. Journal of the American Medical Association 1995 273: 41-5.

British National Formulary, 1995

RC Spencer, PE Wheat, DM Harris. Microcomputer surveillance as an aid to rational antibiotic therapy for urinary tract infection in general practice. Health Trends 1986 18: 84-6.

RA Swann, J Clark. Antibiotic policies - relevance to general practitioner prescribing. Journal of Antimicrobial Chemotherapy 1994 33 (Supp 1): 131-5.

Cholesterol Screening

Dr Andrew Neil (Radcliffe Infirmary, Oxford) wrote to Bandolier 8 about cholesterol screening and the treatment of hypercholesterolaemia.

In Bandolier 5 there was a review of the role of cholesterol screening and the treatment of hypercholesterolaemia. It concluded that screening will not make a contribution to the lowering of overall mortality rates and should be discouraged whilst treatment should be targeted at those patients with the highest overall risk of coronary heart disease. I agree with this, but not with your endorsement of the conclusions of a recently published meta-analysis of cholesterol lowering treatments (Smith, Song & Sheldon, BMJ 1993; 36: 1367-73) which suggested that the likely benefit of treatment was restricted to patients with an annual coronary heart disease mortality rate exceeding 5%.

Implementation of this finding in clinical practice would result in some highly questionable decisions. For instance, lipid lowering drug therapy might be prescribed to men aged more than 80 years since their annual CHD mortality exceeds 3.3%. Drug treatment would be denied to men aged 35-44 years with familial hypercholesterolaemia in whom the annual incidence of fatal coronary heart disease is 1.1%, and the cumulative risk in this condition of a fatal or non fatal myocardial infarction by the age 60 years is about 50%.

The explanation for these anomalous conclusions is that the meta-analysis cited included trials with subjects ranging in age from 18-70 years. Epidemiological principles - and common sense - suggest that the absolute risk of coronary heart disease is unhelpful in determining whether drug treatment is appropriate unless age-specific incidence is taken into account.

Dr Andrew Neil
Radcliffe Infirmary

Hospital-led Prescribing

Bandolier 4 had a short piece on hospital-led prescribing which concluded that there was insufficient evidence to lay the blame for high GP prescribing costs on hospital loss-lead prescribing. This brought strong responses from Dr Tom Jones of the Oxfordshire FHSA and Sharon Hart and colleagues from the Bucks FHSA which express concern that this may undermine efforts to persuade hospitals to co-operate in helping GPs toward more cost-effective prescribing. The following points were printed with the approval of the correspondents:

  • Wiffen & Lauder showed that the cost to the community of a 'basket' of the top 100 drugs by total expenditure was 7% higher than the cost to the hospital pharmacist. This is not insignificant, representing perhaps as much as £250,000 in one quarter in Oxfordshire. In addition, VAT is payable on hospital, but not community drug costs. Taking VAT out of the equation increases the differential above 7%.
  • The comparisons made underestimate the saving potential if alternative drugs were substituted. Thus an Audit Commission Report on prescribing in Buckinghamshire suggested a saving of £220,000 a year if there were full substitution of ibuprofen and naproxen for expensive NSAIDs such as fenbufen and diclofenac. Similar arguments could be used for Co-amilofruse and Co-amilozide being substituted by frusemide and bendrofluazide, and cimetidine for ranitidine.
  • Inevitably much GP prescribing is hospital driven, and rightly so, for hospital specialists are experts in the therapy of the conditions they treat. In most cases medicines are recommended solely for therapeutic supremacy.
  • However, if in even a small number of cases the Hospital Specialist chooses a medicine because of low hospital cost when a therapeutic equivalent would be more expensive but cheaper in the community , the community drug bill goes up unnecessarily. In cases like this, it is difficult for GPs to change prescribing when patients on long-term therapy return to the community. It is important to identify those drugs which do have cheaper community alternatives so that savings overall can be made through enlightened purchasing decisions and rational GP prescribing.
Some hospital staff and managers may take the Wiffen & Lauder article to imply that trying to make savings in the community drug bill through attention to hospital prescribing is unnecessary. This is unfortunate since there are a number of hospital initiatives that could reduce community prescribing costs without loss of effectiveness. The NHSE recently arranged a Prescribing conference for purchasers, in part because of this very point.


Iain Chalmers of the UK Cochrane Centre in Oxford wrote about apparent inconsistency in an article in Bandolier 21 on the presentation of statistical information about the effects of treatment.

I was interested in an apparent inconsistency in the article in Bandolier 21 on the presentation of statistical information about the effects of treatment. Although you conclude that the Number-Needed-to-Treat (NNT) is becoming "the presentation of choice", you suggest earlier in the article that the "Dear Doctor" letter about oral contraceptives would have been more useful if it had expressed the risks associated with them using the "clinically interpretable numbers 1 in 6667 and 1 in 4000".

I agree with you. The NNT is an excellent way of presenting statistical information for those who are considering policies; but I do not think it is as helpful to individual clinicians or recipients of health care as the "Likelihood of Being Helped or Harmed" (LBHH) [1]. Clearly there is scope for some more empirical research here to find out which among the substantial numbers of opinions on these matters - including that expressed above - can be supported by evidence.

Iain Chalmers
UK Cochrane Centre, Oxford.

Bandolier - a ladder for the lottery from Graham Henderson

Whilst I wholly agree with most of the points in your article "Risk" (Bandolier June 1996), and particularly with the need for more widespread understanding of the magnitude of risk, I must take issue with the table (risk ladder) in the article.

As a consistent lottery player who has yet to win his first tenner, I have looked at the statistics and am sure that the risk of winning the lottery (presumably you didn't mean the risk of dying from winning the lottery!) is better than 1:10,000,000-99,999,999.

There are 52 draws per year with, say, 1 jackpot winner per draw (my impression is that there are in fact slightly more). This makes about 52 winners per year out of a UK population of, say, 56 million, giving the risk of winning as just under 1:1,000,000 (level 7). If the calculation is modified to include only those eligible to enter (ie age > 16), or there are more than an average of 1 winner per week, then the risk of winning drops into level 6 (WOW).

Finally, of course, if we include only those exposed to relevant risk factors (ie buying a ticket), the risk changes to a relatively modest 1:13,983,816 per ticket, or 1:269,000 per year for the buyer of a single ticket each week.

If this simply demonstrates that I don't understand the concept of risk myself, please let me know!

Graham Henderson

Verrucas from Dr A R Cadamy , CrossHills, Keighley in response to Bandolier 31 - Verrucas and Games

Dear Sir,

Once again the excitement generated by the verucca question - "To swim or not to swim" has obscured the real issue. Whether or not veruccas are transmitted in swimming pools is irrelevant to the problem of whether sufferers should be excluded from swimming lessons.

To my knowledge, and I confess to not doing a literature search, the mortality rate of veruccas is something approaching zero. Whereas reports of children drowning are an all too frequent occurrence. It becomes a matter of philosophy to choose between the morbidity of veruccas and the mortality of drowning.

There are various approaches to the problem, such as an RCT, to discover the difference in drowning rates amongst verucca-sufferers excluded or not excluded, but they would seem to be fraught with difficulties and impracticalities. Perhaps we first need to assess the NNT (Numbers Needed to Teach) of swimming lessons. How many children do we need to teach to prevent a single drowning? If the figure is very high we could abandon swimming lessons entirely and use the funds more profitably. The incidence of veruccas may or may not then fall but is that not where we came in?

From Dr Dudley C Hubbard Main Practice, 67 City Way, ROCHESTER


I read the article with interest as the treatment and control of verrucae has been an area of endeavour with me for many years. Unfortunately I have no controlled or carefully analysed data to present but over 25 years experience as the main treater of these in our practice of over 10,000 patients. Until the advent of liq. nitrogen available to general practitioners, we used an old Hyfrecator to diathermy the ones that didn't respond to topicals and paring, to which many did respond. Between 1972 and 1992 we used the Hyfrecator about 800 times; we keep a record of cases done. The total number seen and treated must have been several thousand on this basis.

At one stage I had conclusive proof of transmission at one particular school. I became aware of a preponderance of the verruca club from one school. When I checked I found that almost half of the then current group came from one school. We have pupils spread across some seven secondary schools close to the surgery. I felt that this was more than chance and investigated further. I discovered that one particular PE mistress did not believe verrucae were important, nor catching, so insisted all pupils did bare foot gym and dancing on polished wood flooring. No precautions were taken whatsoever.

I insisted that this be stopped, encountered resistance, but with the backing of the county council schools PE inspector, had all active cases covered up and parents' rights to insist on PE in gym shoes re-instated. The epidemic disappeared in about 3 months, returning to the usual sporadic case pattern we see in all the schools.

I cannot prove whether the infection was occurring in the changing rooms or in the gym, but as all schools inevitably have bare foot youngsters in changing rooms, I feel the gym floor was almost certainly the source of cross transfer. The risk of minor trauma from the floor would seem to support the theory that skin abrasions are the entry route.

From: David Jobling , 21 Stockwell Rd, Knaresborough, N Yorks


on Thu, Jan 9, 1997 11:32 am

Subject: Warts!

I wish to take issue with the paper on "Minor surgery - warts and all" in issue 34 . A central assumption is that an average of 5 treatments are required for resolution of the wart. Costings are worked out from this. We audited a years experience of cryosurgery in our practice in Jan. 95. 70% of patients attended for 1 or 2 treatments only. We were encouraged but then wondered if we had scared them all away! We therefore contacted patients who had only attended once. Of the first 14, one had not returned because of the experience, but the other 13 had had their wart healed by a single treatment. This seemed fairly reassuring. Our impression is therefore that an average of less than 2 treatments would be nearer the mark. Also, the authors do not take account of treatments that are given but which are not claimed for because they are in excess of the maximum that can be claimed for under the Minor surgery regulations. These are effectively done for nothing and should reduce the average "cost" per wart still further. Warts do cause much worry to parents particularly in view of the concern in schools about verrucae and treatment with topical paints is fiddly and not very effective - most patients attending for cryotherapy have tried this first. A referral to hospital for treatment locally would cost either £180 per referral, or £60 per attendance - much more than treatment in Primary Care.

From: Stephen Cox , Ivy Court Surgery, Tenterden.

on Wed, Jan 8, 1997 8:55 pm

Subject: Verrucas...

To: andrew.moore

The preventative measure advised by schools and leisure centres is that all verrucas should be cocooned in a rubber sock. The wart virus is widely spread and any new case is merely adding a small addition to the 'viral pool' happily inhabiting the wet and slippery floors of our leisure centres. The logical conclusion is to advise patients who DON'T have verrucas to wear a verruca sock as this will protect them from the trauma that allows entry of the virus in the first place. I floated this idea with our local leisure centre who were unable to follow the thread of my argument. May I propose a nation-wide campaign to make the wearing of verruca socks by the uninfected acceptable? Perhaps with go-faster stripes?

From: Dr Ian Cocks , The Holts Health Centre, Newent, Glos GL18 1BA

More Verrucas and "Games People Play" (Eric Berne)

I was interested in the back page of your December issue about verrucas. I think that the whole issue should be demedicalised and doctors should reduce their control of the problem and treatment should be reserved for people with moderate symptoms. Guidelines in schools should include information about the problem, home management, very little mention of the medical profession!

I don't think there should be any restriction on activities. The logical extension of any restriction would include going back to the crazy verruca slippers that were being sold several years ago which could only have been good for the manufacturers as they only lasted about a week!

Shoulder Pain

from DR ARB Clark, Whitefields Surgery, Hunsbury Hill Road, Camp Hill, Northampton NN4 9UW

in response to Bandolier 32 - Shoulders to the Wheel

Dear Bandolier Editors

As usual I have enjoyed your latest edition of Bandolier and I was interested to read the article about shoulder pain.

Unfortunately I have not audited how many of these I see, but as I am sure that it must be at least one a week. I have found that a substantial proportion of them are due to a problem with the cervical spine. The patient describes the pain across the shoulder, often going a little distance down the upper arm, which usually causes restriction of abduction and may radiate up into the neck.

On examination I often find a painful arc, tenderness and restriction of movement in the cervical spine, usually around C4/5. The patient may have woken with this and my simplistic interpretation is that they have caused some displacement in the joint, due to sleeping on pillows the wrong thickness for their neck.

Having checked that there are no contra-indications, I usually manipulate the joint, by the very simple method of twisting the neck around as far as it can go, and then clicking it the last 2 or 3 degrees of available movement.

In the majority of cases I find that the patient then gets better. I then give them an advice sheet about the care of their neck (copy enclosed for your information) to tell them how not to do it again.

I have seen problems with shoulders that have gone on for many months or a year or more, which have responded to one brief manipulation.

I am not pretending that all shoulder problems are due to this but I see so many that I suspect that it is a common cause of what we see in the surgery.

Unfortunately I have not done a controlled trial or even an audit of my work.

If you clever chaps have any suggestions about what I should do I am willing to consider it!

With my warm greetings to your all.

Yours sincerely

Dr ARB Clark


Your Doctor has given you this advice sheet because you have had problems with your neck. This advise is to help you avoid further injuring your neck.

The commonest cause of neck pain is sleeping with the head incorrectly supported. While asleep the neck muscles become very relaxed and they do not protect the neck joints and discs as effectively as they do while awake. If your pillows are either too thick or too thin it is easy to put out the neck whilst turning over in bed.

The ideal pillow thickness is that which keeps your head in line with your body when you lie on your side. The pillows should fill nicely the gap between your shoulder and your cheek. Even though you will roll over during your sleep, this will still be the correct thickness to support your neck.

Usually the ideal is either one fairly thick firm (foam) pillow or a thin foam pillow with a feather one on top of it.

Wherever you may travel, always make sure that your head and neck are adequately supported before you go to sleep. This particularly applies if sleeping in unusual places such as a car, bus or train etc. Always check your pillows if not sleeping in your usual bed. Never go to sleep with your head propped up on the arm of a sofa.

If you have recurring neck problems it is helpful to do a simple exercise to keep the neck supple. Do a slow, full, head-rolling movement at least five times a day. If you feel dizzy doing it, stop immediately, as sometimes this exercise can affect the blood circulation through the neck.

Alternative Medicine

from Dr Andrew Vickers

in response to Bandolier 31 - Editorial


In Bandolier 31, you claim that there is "pressure to receive alternative medicine into the church of conventional medicine." This choice of language is unfortunate: medicine is not or at least should not be a church; health professionals should rely on evidence, rather than faith. Moreover, it should not be conventional practitioners alone who have the power to determine what therapies are available to patients. A more rational view might be that decisions about health care provision should take place after a broad, evidence-based debate within the community as a whole.

Your point about priorities in health research is a good one. It is worth researching alternative (or "complementary") medicine for four reasons. Firstly, it is used on a large scale by the public. Population based surveys suggest that approximately 10% of the UK population visit a practitioner of complementary medicine each year. Studies of particular patient groups - those with arthritis or HIV for example - often find utilisation rates in excess of 40%. Secondly, there is widescale use by health professionals. In one recent survey of doctors in South Thames, 20% of GPs and 12% of consultants were practising a complementary therapy. Nearly 95% of GPs had referred patients to complementary practitioners. Thirdly, the therapeutic claims made in complementary medicine are interesting and important. If it is true that chiropractic can improve long-term outcome in back pain, that acupuncture is of benefit for migraine, that Chinese herbal medicine can relieve severe, refractory eczema in children or that homoeopathy can reduce the severity of hay fever, then this is surely something that is of interest to practitioners and patients. The final reason for further research is that there is preliminary evidence that at least some claims in complementary medicine do hold. For example, there are good quality randomised trials supporting each of the claims given above.

So has complementary medicine research been given too high or too low a priority? A short mental exercise might be to stop and think of what percentage of the total research budget should be spent on complementary medicine: 1%? 0.2%? 3%? Cost out these percentages given an overall research spend of 2bn per year and it is clear that complementary medicine research is woefully under-funded. The Research Council for Complementary Medicine is an independent charity founded 13 years ago to promote rigorous research on the complementary therapies. Our best estimate is that no more than 3m has been spent on complementary medicine research during this entire period. The NHSE has recently taken this point and has gone at least some of the way towards giving complementary medicine an appropriate place on the research agenda.

Yours sincerely

Andrew Vickers


Vickers A. Use of complementary therapies. BMJ 1994 Oct 29;309(6962):1161

Vecchio PC. Attitudes to alternative medicine by rheumatology outpatient attenders. Journal of Rheumatology. 1994 ;21(1):145-7

Anderson W. O'Connor BB. MacGregor RR. Schwartz JS. Patient use and assessment of conventional and alternative therapies for HIV infection and AIDS. AIDS. 1993;7(4):561-5

Perkin MR. Pearcy RM. Fraser JS. A comparison of the attitudes shown by general practitioners, hospital doctors and medical students towards alternative medicine. Journal of the Royal Society of Medicine. 1994;87(9):523-5

Meade TW, Dyer S, Browne W, Townsend J, Frank AO. Low back pain of mechanical origin: randomised comparison of chiropractic and hospital outpatient treatment. Br Med J 1990 Jun 2;300(6737):1431-7

Vincent CA. A controlled trial of the treatment of migraine by acupuncture. Clin J Pain 1989 Dec;5(4):305-12

Sheehan MP, Atherton DJ. A controlled trial of traditional Chinese medicinal plants in widespread non-exudative atopic eczema. Br J Dermatol 1992;126:179-84

Taylor Reilly D, McSharry C, Taylor MA, Aitchison T. Is homoeopathy a placebo response? Controlled trial of homoeopathic potency, with pollen in hayfever as a model. Lancet 1986 Oct 18;II(8512):881-5

Letter from Dr D R Swinson, Wigan

Department of Rheumatology, Royal Albert Edward Infirmary, Wigan Lane, Wigan WN1 2NN

I share your puzzlement about alternative medicine and I am sure the reason for its popularity has sociological roots and is akin to the continuing faith in Witch Doctors and other traditional healers in the face of the clear effectiveness of 20th Century medicine. There is also a continuing and persistent delight in the paranormal and occult as demonstrated on our television screens. The problem that I have with alternative medicine is that it covers such a wide variety of approaches to medical care from medical herbalism, which is merely conventional 18th Century medicine, to the bizarre claims of homeopathy and reflexology whose belief systems are simply without the boundaries of day to day common sense logic. In this regard I would suggest that Bandolier should simply be discriminatory in the type of alternative medicine which it explores and that it may well be profitable to look closely at the claims for example of medical herbalism as your recent publications have shown.

Best wishes.

Yours sincerely,

D R Swinson

Consultant Rheumatologist

G.P's guide to home birth

Dr R J Anderson, Exeter

Beacon Lane Surgery
109 Beacon Lane
Exeter EX4 8LT

I was very interested to read your article on the G.P's guide to home birth . In your second from last paragraph you mentioned one of the greatest worries for G.P's contemplating home birth is that of the situation where they are faced with a neonatal needing resuscitation. I feel that in your answer you have not addressed it, only to point out that it is a small risk. For example, how much should the GP's agreeing to home births have with them in the way of resuscitation equipment for the neonate and what should their level of training and skill be? Indeed is there any training available for this?

Otherwise I thought your article was well written and quite clear.

Yours sincerely,
Dr R J Anderson

Letter from Dr T J L Malone

The Mill House
2 Palace Mill Farm
Roack Road
Newton Abbot
Devon TG13 0JJ

Dear Sirs,

The use of the word 'Bullshit' ( Bandolier 32 ) is offensive - even to some doctors. Jeeves would never use the term and there is thus a risk of making Bandolier unsuitable bedtime reading.

Yours faithfully,

Dr T J L Malone

Shopfloor epidemiology

Letter from Dr E M Armstrong

BMA House
Tavistock Square
London WC1H 9JP

I was interested to read in the last issue of Dr Down's request for evidence linking symptoms of respiratory disorder with roads and service stations . As you know, the Board of Science and Education reviewed the evidence to date in their 1992 report on Cycling Towards Health and Safety . They are now working on a study of Transport and Health and I therefore asked them to review the current evidence.

The Committee on the Medical Effects of Air Pollutants (CMEAP) reported in 1995 on non-biological pollution, ie the PM 10 particle. They concluded that: 'the reported associations between daily concentrations of particles and acute effects on health principally reflected a real relationship and not some artefact of technique or the effect of some confounding factor'. The studies examined indicated a range of effects from small changes in ventilatory function or exacerbations of asthma through to increases in deaths among the elderly and chronic sick. However, most of the data available is from the US and the committee were concerned regarding their transferability to the UK. The final conclusion of the committee was that: 'there is no evidence that healthy individuals are likely to experience acute effects on health as a result of exposure to concentrations of particles found in ambient air in the UK.

With regard to asthma, the CMEAP conclude that asthma had increased in the UK over the past 30 years but that this was unlikely to be the result of changes in out-door air pollution. The CMEAP report on health effects of exposures to mixtures of air pollutants concluded that there was no evidence that any of the types of pollution episodes commonly seen in the UK caused symptoms or adverse health effects in people who were otherwise well. However, it was thought that adverse health effects may be produced in some persons with chronic respiratory disease and that also there was evidence that some individuals with asthma might experience some deterioration in their condition, although the majority had shown little effect in most studies. Evidence relating to the long term health effects of exposure to the levels of air pollution that occur in the United Kingdom is unfortunately not available

The Government has recently published its consultation document "The United Kingdom Air Quality Strategy" which proposes targets for reducing key pollutants, the primary source of which is motor vehicles. The Government is advised not only by CMEAP but also by WHO guidelines. Within the document there is nothing specific about motorways and motorway service areas and in looking through the literature we have not found any documentation on Dr Downs' specific part of the road transport network. As well as concern with particulate pollution there is also concern about pollution with benzene, which in certain concentrations it is suggested may cause leukaemia. Catalytic converters can reduce the emissions of benzene from exhaust pipes but do not reduce evaporative emissions. Measures to reduce these in vehicles and at petrol filling stations have been the subject of CE directives in 1992 and 1994. The Government has also proposed additional measures including reductions in the aromatic and benzene content of petrol.

I hope that this information is helpful.

Skeins and in-flight decisions ( Bandolier 33 )

From: Andy Chivers on Sun, Nov 17, 1996 10:04 pm

Subject: Geese over Port Meadow

Dear Andrew

I understand, and in fact I think I have seen it happen, that each goose takes it in turn to move to the front, or rather he slips to the back after his spell on the front. But how do they know when to do that?

The process is the same in the peloton of cycle races.

Which leads me to ask-

As well as the beneficial slip streaming effect for the rear cyclist (and goose) is there also a reduction in drag for the front cyclist (or goose) when closelyfollowed by another.

Keep up the good work.


Andy Chivers

From: Martha McQuay

Date: Fri, Nov 15, 1996 14:46


Dear Dad,

I've just been informed by James Smith, who didn't even have to ponder it for a second, that the geese swap over the leader's position while they're flying, and they have a special order in which they swap, so there!

From: Jonathan Kay

Date: Tue, Jan 7, 1997 13:17

Subject: Geese and Bandolier

To: Henry McQuay

You asked how geese choose the leader in the Vee, technically known as a skein.

Leadership rotates between the geese and current research suggests that the pattern and swopping of positions optimises energy expenditure through aerodynamic efficiency.

Schooling of fish appears similar but is much harder to analyse.

A human equivalent exists in cycle road-racing, with the added complexity of co-operation and competition between individuals and teams. The traditional peloton reduces energy expenditure both in the direction of travel, and by angling into cross-winds.

The political analogues are even more intriguing! Just because leadership is variable doesn't mean it is non-hierarchical!


From: A Moloney , Halton General, Runcorn

Date: 30 December

Subject: Skeins of geese

To: Indoor Games Editor

There was an outstanding programme on vortices on the radio in the last couple of years. Each goose flies behind the wing tip of the one in front in such a way that the vortex so produced reduces the effort of flying. They take it in turns to be the one in the front.

Screening blacklist addition ( Bandolier 16 )

From: Brian Budenholzer, MD


I just reviewed your screening blacklist. I reviewed it looking for Bandolier's opinion re: screening for colorectal cancer, especially a Bandolier review of Mandel's study of fecal occult blood testing (FOBT) to screen for colorectal cancer. (Mandel JS, et al. N Engl J Med 1993;328:1365-71.) I did not see mention of colorectal screening, either on your blacklist or your whitelist.

I propose that colorectal cancer screening be added to the blacklist. Endoscopic screening has not been subjected to a randomized trial. Mandel's study of fecal occult blood screening showed identical mortality rates in the screened and unscreened groups. No NNT for all-cause mortality as screening did not reduce (all-cause) mortality. There was a reduction in colorectal mortality of 2.95 per 1,000 persons over 13 years (0.23 per 1,000 per year). As I calculate it, the 95 % CI for the 2.95/1,000 difference in colorectal cancer mortality would be: [error: 0.21/1,000 to 5.69/1,000] (correction: 1.05/1,000 to 4.85/1,000) over 13 years.

The NNT would be 339 over 13 years (95 % CI for the NNT: [error: 176 to 4,762] (correction: 206 to 952) ). I, along with the Canadian Task Force on the Periodic Health Exam, (Ref: Solomon MJ et al. Periodic health examination, 1994 update: 2. Screening strategies for colorectal cancer. Can Med Assoc J 1994;150:1961-70.) am unconvinced that FOBT passes muster as a screening test to be offered to patients.

Thus I propose adding colorectal cancer screening to your screening blacklist, whether by FOBT or endoscopy (sigmoidoscopy or colonoscopy).

Congratulations on the best internet site in the entire medical field. I look forward to each new edition of Bandolier.

Brian Budenholzer, MD
Spokane, WA;USA

Fashion and evidence

Letter from Dr A J Davis , The Surgery, 70a Norwood Road, Southall, Middx UB2 4EY

Dear Dr Moore,

I have been reading Bandolier magazines with interest.

I have however, become concerned by the trend towards evidence-based health care, particularly as it affects general practice. If we only practice medicine as advised by the latest apparently definitive medical articles, we will lose the ability to assess the significance of this article in an historical context. The results of this is that we will ignore at our peril our own clinical experience built up from out personal practice over many years and this will be to the detriment of our patients.

Let me give you an example, in the use of diuretics for management of hypertension. When I qualified about 20 years it was common practice to use thiazide diuretics in combination with slow release potassium, e.g. Navidrex K. Research then demonstrated that there was no point in giving the potassium supplementation as for the vast majority of people this could be obtained satisfactorily from the diet and adding potassium often caused gastric side-effects. we therefore changed out prescriptions to bendrofluazide or hydrochlorothiazide. Further research then suggested that in some patients the serum potassium levels could drop in older patients to a level which would put the myocardium at risk and it was therefore recommended that we change out diuretics to a combination of therapies, together with drugs such as amiloride or triamterene.

Subsequent research showed that these particular diuretics were not without their risk as they could cause hyperkalaemia! Now we are back o bendrofluazide. I expect that in a few years someone will suggest that we should ass potassium supplements!

I would be interested in your comments.

Nasal go-faster strips

Letter from:- Dr Paul D Jackson , Brownstone House Surgery, New Park Street, Devizes, Wilts SN10 1DS

Nasal Go-Faster Stripes (Vol.3 issue 12)

There were several small studies on the effect of external nasal dilators presented at the American College of Sports Medicine Annual Congress in Cincinnati last year. (1,2,3,4). None showed any benefit to athletic performance. Two of the studies (2,3) looked at continuous aerobic activity, the other two studies (1,4) looked at the effect on recovery time from a short burst of anaerobic activity, which equates to repetitive sprint sports (hockey, rugby, football). Two of the trials (2,4) used placebo strips of tape which were as effective as the nasal dilators, the other two trials (1,3) simple compared performance with and without the nasal dilators.

Above a certain threshold of physical activity the nasal airway has a minimal role in ventilating the lungs. One would hope that our top sportsmen might exceed this threshold when representing their country. These nasal strips fulfil one part of the Hippocratic Oath in that they do no harm. They add to the weight of the athlete but this may be offset by a disproportionate increase in their bank balance.

  1. Effect of Breathrite aids on acute anaerobic performance and recovery. Young L. et al . Medicine and Science in Sport and Exercise supplement to Vol 28 No. May 1996
  2. Exercise responses using the Breathe Right external nasal dilator. Huffmann M.S. et al. Medicine and Science in Sport and Exercise supplement to Vol 28 No. May 1996
  3. Effect of the Breathe Right external nasal dilator during light to moderate exercise. Clapp A.J. et al. Medicine and Science in Sport and Exercise supplement to Vol 28 No. May 1996
  4. The effects of an external nasal dilator (Breathe Right) on anaerobic sprint performance. Papenek P.E. et al. Medicine and Science in Sport and Exercise supplement to Vol 28 No. May 1996

Article from Performance Matters sent in by:-

James W Rimmer , Clinical Audit Adviser, Avon Primary Care Audit Group, King Square House, King Square, BRISTOL BS2 8EE

Go faster strips?

It seems that you can't watch any sport on TV now without wondering why so many athletes have cut their noses - they all seem to be wearing little plasters. The answer is that they are seeking to improve performance by using Breathe Right strips. These are used in the hope that by increasing the airflow through the nose, more air can get it the lungs, and therefore more oxygen will be loaded into the blood.

The theory sounds good, however, blood is 99 per cent saturated with oxygen at the lungs anyway, and consequently it seems implausible that a slight dilation of the nasal airway will make any difference. But don't take my word for it. Several research journals have covered investigations into the strips. As you can see below, none of the reports are particularly flattering. However, if you've tried the strips and think that they make a difference, well then, they make a difference!

  1. A study into post-exercise recovery with and without the strips concluded that Breathe Right nasal dilators do not improve the post-exercise ventilatory response: i.e. breathing rate and depth.
  2. Twelve moderately-trained subjects completed seven stages of a 31-minute endurance test, with and without a Breathe Right nasal dilator. Neither ventilation expired nor oxygen consumption was altered by the nasal dilator. The researchers concluded that nasal dilators do not alter the ventilatory response in exercise.
  3. Endurance performance was studied with and without the strips, and it was concluded that Breathe Right nasal dilators do not enhance exercise performance.


  1. Quindry, J.C.; Brown D.D.; Huffman, M.S.; Huffman, M.T. and Thomas, D.Q. (1996). Exercise recovery responses using the Breathe Right nasal dilator. Medicine and Science in Exercise and Sports, 28(5), Supplement abstract 419.
  2. Clapp, A.J. & Bishop, P.A. (1996). Effect of the Breathe Right nasal dilator during light to moderate exercise. Medicine and Science in Exercise and Sports, 28(5), Supplement abstract 525.
  3. Huffman, M.S.; Huffman, M.T.; Brown, D.D., Quindry, J.C. & Thomas, D.Q. (1996). Exercise responses using the Breathe Right external nasal dilator. Medicine and Science in Exercise and Sports, 28(5), Supplement abstract 418.

Letter from Paul Muir , Clinical Audit & Research Department, North Derbyshire Health Care Service, The Lodge, Walton Hospital, Whitecotes Lane, Chesterfield S40 3HN

Re: El Majombero's Poser
I was interested to read your query about go faster stripes.

I have been a runner, cross country Skier etc. for many years and have my own theory about go faster stripes. The theory is based on years of data collection at a variety of sporting events.

The fast bastards always wear the wildest kit!

The more outrageous your kit, e.g. fluorescent lycra, the faster you run, ski or the harder you climb. I have my own go faster 'acid house' running tights which guarantee a personal best, particularly whilst running through built up areas or by groups of teenagers. The fear of ridicule and physical violence improves your speed more effectively than any nose breathing device.

Go faster stripes on the nose will only be effective if they are truly outrageous, perhaps carrying a logo saying 'punch me!' or a similar provocative message.

I would be interested to find other runners or Nordic skiers who would share their sartorial evidence for a bigger evidence base! The findings could be used to improve the performance of our national athletes, ho ho. The call goes out for evidence based lycra fetishism!

From David R Goodwin , Clinical Audit Manager, Royal Victoria Infirmary & Assoc. Hospitals NHS Trust, Queen Victoria Road, Newcastle Upon Tyne NE1 4LP.

Re: El Majombero's Poser

You want evidence so here it is.

  1. Quindry JC, Brown DD, Huffman MS, Huffman MT & Thomas DQ, 1996. Exercise recovery responses using the Breathe Right nasal dilator. Medicine and Science in Exercise and Sport. 28(5), Supplement abstract 419.
  2. Clapp AJ & Bishop PA, 1996. Effect of Breathe Right nasal dilator during light to moderate exercise. Medicine and Science in Exercise and Sport. 28(5), Supplement abstract 525.
  3. Huffman, M.S.; Huffman, M.T.; Brown, D.D., Quindry, J.C. & Thomas, D.Q. 1996. Exercise recovery responses using the Breathe Right nasal dilator. Medicine and Science in Exercise and Sport. 28(5), Supplement abstract 418.

The nub is that these little strips have no effect at all.

These references were taken from Cycling Weekly, 11 Jan 1997.

I'm sure there will be other cyclists sending you the same details.

Fungal Nail Infections - Bandolier Issue 26

Letter from:-
Mr Conrad Jones. Head of Podiatry
Carmarthen & District NHS Trust

Your recent article was restricted to only 2 oral drug regimes for onchomycosis, one of which (griseofulvin) by now should be consigned to the clinically effective dustbin. Additionally, no mention was made of topical agents and professional choice for individual patients. I hope the following is of interest to you.

Having had a series of drug reps to recent staff meetings, and been bombarded with numerous trial reports (and cream cakes), we have now developed a flowchart for choice of treatment. This takes into account clinical presentation, patient's need for treatment, choice of oral or topical agent, and cost. This is set in the context of a podiatrist being able to reduce the bulk of an infected nail, thus making it a) less painful; b) cosmetically more acceptable; and c) more amenable to penetration by a topical agent. We do not routinely send a sample to the Public Health Laboratory at initial assessment, as clinical experience is often sufficient for a podiatrist seeing 80 or so pairs of feet a week, and false negative results are not uncommon.

Our present choices of treatment are topical amorolfine (Loceryl), weekly application, cost £34 for one bottle which should be sufficient for up to 3 nails; or oral itraconazole (Sporanox Pulse), only newly on the market but promises effectiveness in use, fewer days taking tablets and slightly lower cost (£123 vs £132) for total regime than terbinafine. Although the topical regime trials do not claim such a high cure rate as oral, it is preferred by many patients and better results can be achieved in concert with clinical reduction of nail bulk.

Mr Conrad Jones, Head of Podiatry
Carmarthen & District NHS Trust

Fungal Nail Infections - Bandolier Issue 26

Having taken the trouble to write in response to the above issue (copy enclosed), a reply would have perhaps been in order.

experience in communicating with GPs continues to confirm ignorance of optional treatment for a condition which: (a) has a high incidence among the elderly population and (b) is now far more amenable to treatment with the medicaments quoted in my letter of 27 June 1996.

I had hoped that my contribution via your newsletter would have gone some way to improving patient care.

Implementation gap - Bandolier Issue 39

Letter from:-
Dr G A Luzzi
Consultant in Genitourinary/HIV Medicine
Wycombe Hospital, High Wycombe, Bucks

In Bandolier 39 you discuss the importance of the implementation gap between the production of knowledge and its implementation. The idea of setting up implementation task forces would probably be widely supported, but I think the real challenge which faces the NHS is how to fund the processes of change involved.

For example you mention, on viral load tests and protease inhibitors for HIV is a good case in point, with which we have struggled in South Buckinghamshire. The newest interventions and tests are increasingly expensive. Even when there is widespread acceptance of the evidence for their effectiveness and uniform agreement on the need to introduce them, this may be virtually impossible to do in a cash-limited NHS. I believe this is the real issue which requires a task force.

Disinvestment in ineffective treatments and services is often given as a solution - this is not only very difficult to do, but probably wouldn't release enough cash for expensive new developments.

Furthermore, managing change has been made more difficult than ever because of the progressive devolution of financial and managerial responsibility outwards to smaller and smaller units, such as individual clinical services or GP fundholders. This has created a honeycomb of barriers which makes virtually impossible the short term service and financial shifts that are needed to accommodate effective new developments.

The whole structure of the health system needs to be rethought, if we are to develop a truly flexible and responsive system which can meet the real demands imposed by an evidence-based health care service.

Yours faithfully,

Dr G A Luzzi

Letter from a Member of the ENT Sub-Committee of the Comparative Audit Service at the RCS (Eng):-

Peter M Brown, FRCS
Milton Keynes General Hospital
Standing Way
Milton Keynes MK6 5LD

Dear Sirs.

Re: Bandolier number 55 - Evidence for Tonsillectomy

Many of us are look to publications such as Bandolier to provide us with solid evidence based data. It is therefore very disappointing to find that the quality of your evidence based analysis can be so poor as to be frankly misleading.

The basic problem is that of reporting evidence from a single and rather poor systematic review by Marshall and then throwing doubt on the surgery without adequate evidence.

I was suspicious your knowledge must be poor when I read both Bandolier and Marshall's introductions claiming that the indication for tonsillectomy is either "tonsillitis" or "pharyngitis". No ENT surgeon I know of performs tonsillectomy for pharyngitis, which in the general population is much more common than tonsillitis. I suspect that the reason for this is that GPs may make a different diagnosis than the consultant and North American practice (which is quoted extensively in the paper) may be very different from the UK.

The most important trial in the UK was the Scottish Tonsillectomy Audit (OK not an RCT, but a very extensive audit of practice and of patient's perception of benefits or failure of surgery) which was published in 1996 and although cited by Marshall, is only mentioned by him in passing. He prefers to quote from papers which have at maximum 251 patients, while the Scottish Tonsillectomy Audit had 9,773 patients, of which 3,385 were inpatients. I suppose Marshall did not want to quote from this extensive review of outcome because it did not fit in with his theory, ie, that tonsillectomy was a useless operation.

He would therefore not have wanted to have quoted from this article that recurrent tonsillitis was by far the most common reason for tonsillectomy, but that other significant reasons were upper respiratory tract obstruction caused by tonsils. Also that for people having tonsillectomy the median number of episodes of tonsillitis per year was 5 or 6 with most people being incapacitated for 3 to 7 days per episode. The 12 month follow up showed that 80% of patients after tonsillectomy reported that they had lost no time off work or school since surgery, and that only 2% reported that the operation had not helped at all. Over 90% of patients reported that they were better of "cured". The patient questionnaire in the Scottish audit had an area for free text and the following quote is typical - "they felt they had to fight their GP to obtain a referral to hospital and sometimes had to overcome an apparent resistance on the part of the surgeon to list them for tonsillectomy. 'The operation should have been done years ago' was a commonly expressed view".

The evidence provided by our colleagues north of the border has just been confirmed by the Comparative Audit Service of the Royal College of Surgeons of England, of which I am a member. Although we do not yet have the completed final analysis, the following stats are unlikely to change much. 2,000 patients approximately were in the study. The most common reason for tonsillectomy was by far recurrent tonsillitis, followed by upper airway obstruction. Pharyngitis was not even mentioned as a reason. 74% patients had suffered more than 4 weeks illness in the previous 12 months. The patient questionnaires were returned to the College for analysis, not the surgeon or hospital where they had their operation. 92% said their throat symptoms were better than before, and 72% that their general health was better. Only 1.6% and 2.8% respectively reported a deterioration in both outcomes.

The operation is one of the most painful we perform on either children or adults, yet 86% told the independent College assessors they were glad they had had the operation - and the patient's opinion is far more important than that of any doctor, clinician or otherwise, don't you think?

It takes a long time to build credibility, and only a moment to lose it. Take more care with your quality control.

Letter from:-

Peter J Robb, BSc MB FRCS FRCSEd
Consultant Otolaryngologist
Surrey RH3 7HJ

Dear Sirs,

I wish to comment on two articles which recently appeared in Bandolier.

Tonsillectomy for sore throats : Vol 5, Issue 9, September 1998

It is generally accepted by ENT specialists that tonsillectomy is indicated primarily for recurrent acute tonsillitis, typically occurring three to four times per year, for two years of more. Sore throats, per se, usually of viral origin, will generally not respond to treatment by tonsillectomy.

An analysis of parental ability to differentiate sore throats from tonsillitis by symptom clustering, showed a high level of accuracy and reliability. (Capper R, Canter RJ, Robinson A, 1998. British Association for Paediatric Otolaryngology Annual Meeting.)

It is of note, that currently, approximately 20% of tonsillectomies are performed for upper airway obstruction in children of an age group typically much younger than those experiencing recurrent acute tonsillitis.

There is no evidence that I am aware of, which substantiates the claim that laser surgery reduces secondary haemorrhage rates. Blood loss at the time of operation may be less with laser tonsillectomy, but this is also true of electrocautery removal of the tonsils. There is published empirical evidence that laser removal of tonsils, (KTP or carbon dioxide), produces much deeper and more lateral tissue damage than other methods, and therefore, more severe and protracted post-operative pain.

The routine use of preoperative intravenous hydration, ondansetron and non-narcotic analgesics combined with a rapidly reversing inhalational agent, (e.g. Sevoflurane), has almost completely abolished post-operative nausea and vomiting in children and adults undergoing tonsillectomy.

Post-operative infection is overdiagnosed both by parent and non-ENT clinicians and nurses. The healing, by secondary intention, in the pharynx of the tonsil beds, produces a white slough which is frequently interpreted as infection. In the absence of foeter or fever, this is unlikely, but often treated with antibiotics by general practitioners under pressure from patients or parents.

Having made these observations, I would entirely agree that a decision to undertake, or indeed, undergo tonsillectomy should never be taken lightly. Most ENT surgeons, including myself, spend a good deal of time pre-operatively counselling patients and parents of the relative risks and merits of the procedure.

Despite the two weeks of misery that generally follows a tonsillectomy, the vast majority of parents and patients report an overwhelming positive view of the outcome once the convalescence is over.

Acute otitis media October 1998

Many ENT surgeons with a special interest in paediatric ENT, including myself, share the anxiety that a sea-change away for antibiotic treatment for otitis media will ultimately result in a resurgence of the serious and sometimes fatal complications of this common condition which were commonplace in the pre-antibiotic era.

Most children referred to an ENT clinic are suffering an attack as frequently as every 4-6 weeks, and from a specialist perspective, the 7-year old who has had one ear infection does not exist!

My own research indicates that many pre-school children with severe recurrent acute otitis media have a partial, selective IgA deficiency which is maturational. This occurs in up to 1:700 normal children. The majority seem to mature, and the infections resolve by about the age of four years. These children frequently respond to low-dose prophylactic antibiotic treatment, particularly over the winter months.

It is unclear from your summary of the meta-analysis, whether the management of acute otitis media with amoxycillin 750 mg bd for two days was included as a treatment option. This would be a contemporary treatment for the condition with the benefit of high parental compliance. Azithromycin for three days has the same advantage for those allergic to penicillin.

From the above, you will understand that the associated management decisions of two seemingly simple and common conditions are complex, and one is hesitant to apply or support a broad-brush approach to clinical decision-making advice to non-specialist colleagues.

From: "Rachel E Jordan" < >
Subject: Our correspondence re psoriasis & eczema review
Date: Thu, Apr 8, 1999, 10:50 am

The Editors
Pain Relief Unit
The Churchill

29th January, 1999

Dear Sirs,

Treatments for psoriasis
We are writing with reference to your Bandolier article of May 1998(1) which appraises a systematic review of psoriasis treatments(2). Prompted by your recommendation, we attempted to use the psoriasis review to inform a commissioning decision on the use of cyclosporin A. As a consequence, we made our own appraisal and have to report that we came to rather different conclusions. Although we agree that the search for studies was rigorous and worthy of commendation, the value of the work was greatly reduced because the authors:

(a) combined data without assessing whether it was appropriate to do so, particularly by investigating the homogeneity of results. They combined data from studies with different populations, different measures of psoriasis severity and different outcomes. Despite having over 43 randomised controlled trials, they combined data from both randomised controlled trials and non-randomised uncontrolled trials.
(b) used inappropriate methods to combine the results.
(c) did not give any details of the primary studies which would enable the readers to make correct calculations themselves.

Bandolier's appraisal brought out some of these important points, but despite the recognised and serious limitations of the analyses, printed details of some of the results and still went along with one of the original authors' key conclusions regarding the suitability of this review as the basis of treatment guidelines. We would strongly disagree with this conclusion because the results are not useful, rather are misleading. Our views are supported by the correspondence in the British Journal of Dermatology(3) regarding the original article.

In our opinion, this review promised much, but is of limited use unless reanalysed.

Yours faithfully,

Rachel Jordan Chris Hyde

Department of Public Health and Epidemiology
University of Birmingham
1. Treatments for psoriasis. Bandolier May 1998; 51(5) issue 5:6
2. Spuls PI. A systematic review of five systemic treatments for severe psoriasis. British Journal of Dermatology 1997; 137: 943-9
3. Williams, HC. Li Wan Po, A. Murrel, D. Naldi, L. Diepgen, T. A systematic review of five treatments for severe psoriasis [Letter]. British Journal of Dermatology 1998; 139:757

rachel & chris
I had your letter this morning, and was a bit taken aback, because my recollection was that in the Bandolier article I had made the point about bias in this review. When I read what was written, this to me seemed to be the case:
"This type of analysis is clearly subject to bias, since it includes non-randomised studies. The authors point out their difficulty in ensuring that patients had similar degrees of severity of psoriasis. They also point out that the clear differences between phototherapy and oral treatments reflect the difference in treatment goals. With the former it is full remission, while with the latter it is to induce remission while keeping adverse effects to the minimum by using the lowest possible oral dose."
These aren't weasel words, but describe the problems with the review. Most reviews set out to give us an answer, but they also often set out problems, intentionally or not, because they raise issues about patient selection, treatment regimens, and/or outcomes measured. But they do (or should) give us the baseline - that is a systematically searched set of papers. When I wrote that this could form the basis of treatment guidelines, it was my intention to convey that the set of papers identified in the review was the basis on which guidelines would be established by analysing them according to whatever criteria was deemed necessary. I did not read it then, nor do I now, as suggesting that the results themselves should stand as the basis for guidelines. Could it be read like that, as you have done. I would have thought not, because of all the problems identified in the previous paragraph. But the fact that you have done so must prove me wrong. Even so, I cannot perceive that anyone in their right minds would use the crude results, and pointing to a review which, though wrong in its execution, contained the raw material for sensible work to be done seemed then, and seems now, to be a useful thing to do. Bandolier is a signpost, not a destination. While we do mini-reviews from time-to-time, we do not have the resources to do all the work. Showing others where the raw material are is part of our remit. I am happy to put your letter into the Bandolier correspondence columns, with this reply. Please send me an electronic version, if you can.
With best wishes
Dear Andrew,
Thanks for your prompt response to our letter and for your clear explanantion of Bandolier's portrayal of this piece of work. However, the serious point is this. We both know that most people rarely read the substance of an article, which is where your provisos about this review are made, and as a result it is highly likely that the only message a casual reader will take away from your article is "This review could form the basis of treatment guidelines." This would be true if the authors had given us some information about the studies they had included (and excluded) but they don't. They don't even make the offer that a full list of included studies would be available from them on request. Without access to such a resource I am completely at a loss as to how this article could be turned into a guideline. Further, the graphics you present give a completely erroneus impression that the responses are comparable in a way which even the original article does not do. We don't wish to create artificial conflict just fill correspondence columns, but I do think it is worth making the point that we all need to be very careful about the bottom-lines we supply people whether it be in the form of conclusions in abstracts, "parting shots" or powerful graphical presentations of results. I would be happy to redraft our article to raise this key point, using the psoriasis review as a particular example if you think this would be a more constructive approach, rather than just apparently being critical of Bandolier editorial policy - which was not our main intention; I still have a great deal of respect for what you do, and suspect that you contiune to do it under heavy constraints. Finally we clearly think the Bandoiler portrayal of this article was misleading, where as you don't. I would argue that we are probably in a better position to make that judgement. However, the truth of the matter is easily tested. Why don't you ask the Bandolier readers what they took away from this article - did they use it as a sign-post to get the original article; or did they just take the comparison of the treatments at face value. Why don't we do the above and then publish correspondence as appropriate to the results. Chris & Rachel.
Rachel Jordan
Dept Public Health & Epidemiology
University of Birmingham
Birmingham B15 2TT
tel 0121 414 6775
fax 0121 414 7878

Hospital acquired infection Bandolier 73, February 2000

It's difficult to know whether to laugh or cry when the NHS publishes such

Work (Hospital acquired infection, Bandolier 73).

I'm sure the data are correct but there are two MAJOR flaws in the conclusion that infection causes death!

1) Lower respiratory tract infections (probably the most important single group of infections) are caused by organisms migrating from the upper airways/larynx/pharynx and are largely unrelated to hand washing or infection control procedures. In ICU where stringent cross infection controls are in place patients still develop pneumonia due to their own micro-organisms which may be selected out by antibiotic use.

2) Patients who develop nosocomial infections probably represent an "at risk" group either because of environmental/interventional factors or due to a genetic predisposition to infection or adverse outcome. Unless the authors are able to control for these factors they cannot justify their conclusions. In the ICU, where we are able to stratify for severity and risk there still remains the question - do patients die from ICU acquired infection or are infections merely a marker of patients who are destined to die?

Best wishes and continue the good work.

Christopher S Garrard

Intensive Care Unit

John Radcliffe Hospital

Oxford OX3 9DU