Authorship. Team approach to assigning authorship order is recommended

Gulf war illness

Why it took so long to decide toinvestigate

Editor—As it is now six years since the Gulfwar, it is reasonable to ask why the problemof Gulf war syndrome has not beenresolved.1 As surgeon general in 1991-4 Iwant to comment on the unravelling time-table during these six years.

At no time have we denied that Gulf vet-erans have been ill—indeed, whenever possi-ble we have actively recommended theirappropriate medical management. Sec-ondly, despite press reports, we have alwayskept an open mind on the presence of aspecific Gulf war syndrome.

The Gulf conflict in 1990-1 itself wasshort, casualties mercifully few, and the sick-ness rates unremarkable; likewise, whentroops returned to the UK the daily sicknessrate did not increase. Also at that stage littleevidence existed of psychiatric illness orpost- traumatic stress syndrome.

Some 18 to 24 months later, however, webecame aware of a campaign by lawyers torecognise a specific Gulf illness. Up to 200people were on their books, but despitewidespread appeals for them to comeforward, including two television appeals,only 15 approached the armed forces medi-cal services. Not surprisingly, among these15 no pattern of illness emerged. They wereall managed appropriately.

As we were keen to solve the problem,in October 1993 we appointed a singleconsultant physician, Group Captain Coker,to see all such patients. At first only a trickleof patients came forward—even by June1994 (three and half years later) he had seenonly 30 patients. Over the next two years,however, these numbers increased so that byJanuary 1997, 1100 had been registered, 920had been seen and fully investigated, andresults had been completed and analysed in519.

A preliminary paper in the summer of1996 concluded: “From our small databaseit would not appear that there is a uniquecluster of symptoms which would suggest anew pattern of illness in veterans.”2 Therewas, however, a higher than normal inci-dence of psychiatric and stress relatedillness. A large battery of tests was uniformlyperformed on all the patients, with furtherinvestigations as necessary.

Now it has been agreed that fullepidemiological studies are to be under-taken into the incidence of ill health among

UK Gulf war veterans and into theirreproductive health. In answering thecriticism that these expensive studies shouldhave been instigated earlier, I have tried toshow how difficult it has been to decideexactly when this decision could have beenmade. Early on there were very few patientsand no evidence of a syndrome. At all stageswe have liaised with our counterparts in theUnited States. They embarked a little earlieron similar studies as they had similar resultsto their investigations but dealt with greaternumbers.

An important reason for writing this let-ter is because our armed forces may well beplaced in similar alien environments infuture and we owe it to them to provide thebest medical support, which of courseincludes planned protective and scientifi-cally justified measures against likely threatsto their health.Peter Beale Chief medical adviserBritish Red Cross Society, London SW1X 7EJ

1 David A, Ferry S, Wessely S. Gulf war syndrome. BMJ1997;314:239-40. (25 January.)

2 Coker WJ. A review of Gulf war illnesss. Royal NavalMedical Service Journal 1996;82:2.

Family doctors should be aware ofassessment programme

Editor—Your editorial by David et al onGulf war illness1 was prompted by the recentpublication in JAMA of four papers on thistopic. Rightly, the authors point out thatthese papers provide more questions thananswers and are “hypothesis generatingrather than hypothesis testing.”

Yet surely the studies they cite indicatethat there is too much smoke in this area toexclude the existence of a genuine fire; inother words, the possibility remains thatsome veterans of this conflict have becomeill directly as a consequence of their servicein the Gulf war.

That more research is needed isgenerally agreed, and we welcome theprojects recently approved by the MRC andfunded by the Ministry of Defence, eventhough it has taken over five years to reachthis point. The Royal British Legion believesthat not all veterans of the Gulf war who nowhave health problems are aware of theopportunities open to them. There exists athorough medical assessment programmeorganised by the ministry. We write prima-rily to draw the attention to family doctors tothis programme. Veterans who are unwellmay be referred to Lt Col Bhatt, Central

Medical Establishment, Kelvin House, 32-34Cleveland Street, London W1P 6AU (0171636 4651, ext 211). In a recent letter to thelegion the assistant private secretary to theministry has undertaken to provide travelwarrants and, where necessary, overnightaccommodation for veterans travelling toLondon to take advantage of this pro-gramme. Details are available from thelegion.Peter Heaf Honorary medical consultantNorman Jones Honorary medical adviserRoyal British Legion, London SW1Y 5JY

1 David A, Ferry S, Wessely S. Gulf war illness. BMJ 1997;314:239-40. (25 January.)

Chronic forearm pain

Thermography is a sensitive adjunct todiagnosis

Editor—Reporting on thermographicchanges in keyboard operators with chronicforearm pain, S D Sharma and colleaguessuggest that thermography needs furtherevaluation as a diagnostic tool and that itmay become a useful means of assessing theefficacy of treatment.1 Our experience over afive year period of the use of thermographyas an aid to diagnosis and later as a means of

Advice to authorsWe receive more letters than we can publish: wecan currently accept only about one third. Weprefer short letters that relate to articlespublished within the past four weeks. We alsopublish some “out of the blue” letters, whichusually relate to matters of public policy.

When deciding which letters to publish wefavour originality, assertions supported by dataor by citation, and a clear prose style. Lettersshould have fewer than 400 words (please give aword count) and no more than five references(including one to the BMJ article to which theyrelate); references should be in the Vancouverstyle. We welcome pictures.

Letters should be typed and signed by eachauthor, and each author’ s current appointmentand address should be stated. We encourage youto declare any conflict of interest. Please enclose astamped addressed envelope if you would like toknow whether your letter has been accepted orrejected.

We may post some letters submitted to us onthe world wide web before we decide onpublication in the paper version. We will assumethat correspondents consent to this unless theyspecifically say no.

Letters will be edited and may be shortened.

Letters

1041BMJ VOLUME 314 5 APRIL 1997

objective assessment of the effects oftreatment on horses in a veterinary practicestrongly supports this view.2

The initial subjects were 49 horsessuffering from persistent non-specific lame-ness or chronic back pain which had notresponded to standard veterinary treatment.All were referred for manipulative treatmentusing osteopathic techniques. More recentlysome 400 horses have been assessed beforeand after treatment, and in 80% of equinecases with chronic back pain, no orthopae-dic or neurological pathology was found.3

The parallel with back pain in humans is atonce apparent.

Like Sharma and colleagues, we wereable to show that many subjects had areas ofdistinct cooling in the affected limb, or inour cases often over the paravertebralmusculature. Thermography after treatmentshowed resolution or partial resolution ofthese areas of cooling where there was clini-cal improvement.

Thermographic changes correlate wellwith radiographic changes.4 However, in ourexperience thermography is useful indetecting more subtle forms of dysfunctionwhich often seem to be related to difficultclinical areas currently at the edge of, orbeyond, present concepts of disease. Manyof these require a terminology containingprefixes such as “non-specific,” "atypical,”"primary,” and “functional.”

Although much work is needed instandardising and interpreting thermalimages, thermography can provide a sensi-tive, non-invasive, and relatively inexpensiveadjunct to the diagnostic process, and anobjective means of assessing the efficacy of arange of treatments.C M Colles Veterinary surgeonAvonvale Veterinary Group, Ratley, BanburyOX1 56DFA G Pusey Registered osteopathJ Brooks Registered osteopathA G Pusey and Associates, Haywards Heath, Sussex

1 Sharma SD, Smith EM, Hazleman BL Jenner JR.Thermographic changes in keyboard operators withchronic forearm pain. BMJ 1997;314:118. (11 January.)

2 Colles C, Holah G, Pusey A. Thermal imaging as an aid tothe diagnosis of back pain in the horse. In: Ammer K,Ring EFJ, eds. Proceedings of the 6th European congress ofthermography, Bath 1994. Vienna: Uhlen Verlag, 1995:164-7.

3 Jeffcott LB. Disorders of the thoracolumbar spine of thehorse—a survey of 443 cases. Equine Vet J 1980:12:197-210.

4 Weinstein SA, Weinstein G. Thermography and otheraccepted diagnostic modalities: a comparison. J Osteo-pathic Med 1988 Jan:33-6.

Repetitive strain injury is an unhelpfulterm

Editor—I find it distressing that the BMJ canpublish an article which accepts and seeks topropagate the term repetitive strain injury.1

S D Sharma and colleagues show thatpatients with undiagnosed pain exacerbatedby work had alterations in their skin bloodflow after performing a task which madetheir pain worse. It is well accepted that paincan cause changes in skin blood flow andthereby in temperature. Exactly the sameeffects would be found in a series of patientswith osteoarthritis of the knee asked to use atreadmill. The study simply shows that these

patients have arm pain that is made worseby exercise; it does not prove the existenceof repetitive strain injury. The use of theterm is misleading and unhelpful to patientsand those who treat them. The BritishOrthopaedic Association’s working partywas clear that there was no evidence thatrepetitive strain injury was a real disease,2

and this has been confirmed in a recent edi-torial by the council of the American Societyfor Surgery of the Hand and an editorial inthe BMJ.3 4

In a series of 70 patients with a present-ing diagnosis of repetitive strain injury, Ihave been unable to make a firm medicaldiagnosis on only five occasions. Thediagnosis of repetitive strain injury pro-foundly affects that patient’s future pros-pects, both of recovery and future employ-ment. I have had several patients withclearcut clinical diagnoses who have toldme, “I cannot get better, I have RSI.” Theyhave then declined treatment and left theclinic. A doctor would do much better toexplain to the patient that they have armpain for which present medical knowledgecan supply no diagnosis. Referral to a painclinic is usually then appropriate for painmanagement as opposed to disease treat-ment.

Patients attending my clinic have usuallybeen off work for several months, andclinical signs can be hard to elicit. We there-fore send all patients for a controlledexercise programme to get the arm backinto use. This will temporarily increase thepain but will allow clinical signs to show andtreatment to be started. Simultaneouslypatients are referred to the pain consultantfor management of symptoms until treat-ment is organised. Patients with what I term“complex upper limb pain” should not begiven the dangerous diagnosis of repetitivestrain injury. Rather than being seen in gen-eral orthopaedic or rheumatological clinicsthey should be seen in specialist clinics,where the facilities described above areavailable and which have a high success ratein diagnosing and treating such patients.L C Bainbridge Consultant hand surgeonPulvertaft Hand Centre, Derbyshire RoyalInfirmary, Derby DE1 2QY

1 Sharma SD, Smith EM, Hazleman BL, Jenner JR.Thermographic changes inkeyboard operators withchronic forearm pain. BMJ 1997;314:118. (11 January.)

2 Barton NJ, Hooper G, Noble J, Steel WM. Report of BOAWorking Party on occupational causes of upper limb disorders.London: British Orthopaedic Association, 1990.

3 Weiland AJ. Repetitive strain injuries and cumulativetrauma disorders. J Hand Surg 1996;21A:337.

4 Brooke P. Repetitive strain injury. BMJ 1993;307:1298.

Disorders of spermatogenesisin Finland

Is this a period effect, and if so, why?

Editor—Jarkko Pajarinen and colleaguesreport a dramatic deterioration in sperma-togenesis between 1981 and 1991 atnecropsy in middle aged Finnish men whodied suddenly.1 The rate of abnormalityseems extremely high, even in 1981. Could

this be explained by, for example, the differ-ent fixatives used or by varying delaybetween death and necropsy? Have similarfindings been reported in other comparablestudies?

The authors discuss the discrepancybetween their findings and the evidence thatsperm concentration has not declined inFinland up to as recently as 1994.2 Theexplanation given, that blocked seminifer-ous tubules together with normal sperma-togenesis can coexist with an impairedsperm count, cannot account for observa-tions that are the other way around. Whatcould be relevant, however, are differences inage distribution and in location withinFinland.

Assuming that these findings can beaccepted as real, a key question is whetherthey are due to a cohort effect or a periodeffect. A cohort effect corresponds to anexposure in early life that led to apermanent defect. A period effect resultsfrom an acute or chronic exposure thatoccurred, or at least had its effect, between1981 and 1991. Pajarinen and colleaguescould have investigated this by analysingeach of their cross sectional studies by age(allowing for a possible “normal” back-ground deterioration): a cohort effect wouldpredict that older men had better sperma-togenesis, whereas a chronic period effectwould predict the reverse as older menwould have been exposed for more years. Ifboth effects were present they would tend tocounterbalance one another. An acuteperiod effect would predict no relation withage, unless susceptibility varied with age.

On the face of it, the steep rise inarrested spermatogenesis would suggest anacute period effect—that is, a sudden eventthat occurred between the two studies. Onepossibility is the Chernobyl accident of1986, during which Finland received anappreciable dose of radiation; the peakradiation dose in southern Finland occurredsomewhat to the east of Helsinki.3 A slightrise in congenital malformations showing adose-response relation4 was reported in Fin-land after the accident.3 Although the exter-nal dose was small, certain substances, suchas plutonium and indium, are actively takenup by the adult testis5 and could causeappreciable damage by emitting highenergy radiation.Michael Joffe Senior lecturer in public healthImperial College School of Medicine at St Mary’s,Norfolk Place, London W2 1PG

1 Pajarinen J, Laippala P, Penttila A, Karhunen PJ.Incidence of disorders of spermatogenesis in middleaged Finnish men, 1981-91: two necropsy series. BMJ1997;314:13-8. (4 January.)

2 Vierula M, Niemi M, Keiski A, Saaranen M, Saarikoski S,Suominen J. High and unchanged sperm counts of Finn-ish men. Int J Androl 1996;19:11-7.

3 Harjuleto T, Aro T, Rytomaa T, Saxen L. The accident atChernobyl and outcome of pregnancy in Finland. BMJ1989;298:995-97.

4 Joffe M. The accident at Chernobyl and outcome of preg-nancy in Finland. BMJ 1989;298:1384.

5 Hoyes KP, Sharma HL, Jackson H. Hendry JH, Morris ID.Spermatogenic and mutagenic damage after paternalexposure to systemic indium-114m. Radiat Res 1994;139:185-93.

Letters

1042 BMJ VOLUME 314 5 APRIL 1997

Author’s reply

Editor—Michael Joffe presents an interest-ing hypothesis of a possible associationbetween deteriorating spermatogenesis inFinland and the Chernobyl accident in1986. We do not think that radiation couldentirely explain our finding, however, asmost of our necropsy data came from theHelskini area, which avoided the peak ofradiation. The hypothesis does, however,raise new questions which should be consid-ered in future investigations.

The incidence of normal spermatogen-esis (56%) in our series correlates well withanother study which reported normalspermatogenesis in 50% of cases.1 Inaddition, our series included heavy drinkers,in whom there is a high incidence ofdisorders of spermatogenesis.1-3 This signifi-cantly decreased the mean incidence of nor-mal spermatogenesis in our series. Interest-ingly, when moderate drinkers wereanalysed in the 1981 series normal sperma-togenesis was found in four out of five men.The role of alcohol consumption in thedeterioration of spermatogenesis was not,however, included in our study.

To allow for confounding from the vari-ous types of fixation methods we assessedthe effects of certain fixatives on sperma-togenesis in a previous study.2 Formalincaused detachment of germinal epitheliumin seminiferous tubules, making the scoringof spermatogenesis to some extent moredifficult. Neither formalin nor Bouin’s solu-tion had an effect on the state ofspermatogenesis. Recently, we investigatedthe effects of fixation on testes with an iden-tical state of spermatogenesis. No associ-ation was observed between testicularmorphology and the type of fixativesolution (unpublished data). In addition, wedo not think that the slight difference of 0.3days (3.5 in 1981 v 3.8 in 1991) betweendeath and necropsy could explain thedifference between the two series; reasonsother than technical might explain thisfinding.

As we pointed out in our article,blockage of seminiferous tubules has beenobserved in several oligospermic men,suggesting an inconsistency between semenanalysis and state of spermatogenesis.4 Thetotal sperm reserve is unlikely to becompletely depleted after one ejaculation.5

A slight deterioration in spermatogenesis,such as partial spermatogenic arrest, couldproduce enough spermatozoa to make acomparison with normal spermatogenesisdifficult, especially if analysis was performedafter a period of abstinence, during whichsperm reserves become replenished. Ifsemen analysis was performed at morefrequent intervals differences in sperma-togenesis might become more apparent.Jarkko Pajarinen DoctorDepartment of Forensic Medicine, PO Box 40,SF-00014, University of Helsinki, FinlandPekka J Karhunen ProfessorSchool of Medicine, University of Tampere, Finland

1 Kuller LH, May SJ, Perper JA. The relationship betweenalcohol, liver disease and testicular pathology. Am J Epide-miol 1978;108:192-9.

2 Pajarinen JT, Karhunen PJ. Spermatogenic arrest andSertoli cell-only syndrome—common alcohol-induceddisorders of the human testis. Int J Androl 1994;17:292-9.

3 Boiesen PT, Lindholm J, Hagen C, Bahnsen M, Fabricius-Bjerre N. Histological changes in testicular biopsies fromchronic alcoholics with and without liver disease. ActaPathol Microbiol Scand 1979;87:139-42.

4 Meinhard E, McRae CU, Chisholm GD. Testicular biopsyin evaluation of male infertility. BMJ 1973;iii:577-81.

5 Oldereid NB, Gordeladze JO, Kirkhus B, Purvis K.Human sperm characteristics during frequent ejacula-tion. J Reprod Fertil 1984;71:135-40.

Penetrating intraoral traumamay occur in adults as well aschildrenEditor—R C Law and colleagues present anexcellent account of the hidden dangers ofpenetrating intraoral trauma in children.1

Although this type of injury is morecommonly seen in children,2 it should not beoverlooked in adults.

We recently treated a 28 year old manwho had been assaulted with an umbrella.An obvious lip injury was noted, but anoropharyngeal injury, in the same trajectory,was missed, and he presented 24 hours laterwith hemiplegia and aphasia. Contrast com-puted tomography and angiography con-firmed thrombosis of the carotid arterysecondary to trauma to the posteriorpharyngeal wall (fig 1). Thrombosis can alsooccur secondary to blunt head and necktrauma.3

We would re-emphasise the need tomake parents aware of the dangers ofchildren walking around with sharp objectsin their mouth. We would also recommendthorough oral examination in patients—adults as well as children—presenting withperioral as well as intraoral trauma.S S Rayatt Senior house officer in plastic surgeryP Hamlyn Consultant in neurosurgeryRoyal London Hospital, London E1 2BBP Magennis Specialist registrar in maxillofacialsurgeryWalton Hospital, Liverpool L9 1AE

1 Law RC, Fouque CA, Waddell A, Cusick E. Penetratingintra-oral trauma in children. BMJ 1997;314:50-1.(4 January.)

2 Hengerer AS, Degroot TR, Rivers RJ, Pettee DS. Internalcarotid artery thrombosis following soft palate injuries: acase report and review of 16 cases. Laryngoscope1984;94:1571-5.

3 Pretre R, Reverdin A, Kalonji T, Faidutti B. Blunt carotidartery injury: difficult therapeutic approaches for anunderrecognised entity. Surgery 1994;115:375-81.

Suspension of nurse who gavedrug on consultant’sinstructions

Consultant is grateful for readers’support

Editor—I wish to thank the many readerswho have supported my action and that ofthe sister in a day hospital in giving apsychotic patient surreptitious drugtreatment.1 Needless to say, this was verymuch a last resort. I did not give the patienthis drugged tea myself because he would nothave accepted it from me; the nurseconcerned naturally hesitated when I askedher to give it, and I foolishly assured her thatI would take full responsibility.

Since then I have done everything in mypower to persuade the hospital administra-tion to put matters right—without success. Ihoped that publicity might facilitate this.The chairman of the trust, however, has toldme that it is up to the chief executive, whilethe chief executive has told me that she willtake no action. Meanwhile, the senior nursewho initiated the disciplinary action hasbeen promoted and has left, while the sisterin the day hospital continues to give devotedservice to her patients without complaint,though she is deeply hurt by the treatmentshe received.John M Kellett Consultant geriatricianSt Georges’s Hospital Medical School, University ofLondon, London SW17 0RE

1 Suspension of nurse who gave drug on consultant’sinstructions [letters]. BMJ 1997;314:299-301. (25 January.)

Patients should be alert to doctors’willingness to use subterfuge to givedrugs

Editor—I am appalled to learn that asubstantial number of doctors believe that itis sometimes acceptable to use subterfuge togive drugs to cognitively intact patients whohave not been detained under the MentalHealth Act without their consent and thatsome of these doctors are prepared to puttheir beliefs into practice.1 2 I had hithertoassumed that patients who claimed that suchthings were going on were paranoid, but Iwonder now if on occasion my attempts toreassure them were misplaced. Until theGeneral Medical Council sees fit to con-demn this behaviour, I suggest that it shouldissue the following advice to the generalpublic:

“In certain circumstances a doctor maydecide to give you a treatment that you havenot consented to, or even one that you havespecifically refused. He or she may do this byconcealing a drug in your food or drink, or

Fig 1 Intravenous digital subtraction angiogramarch aortagram showing complete occlusion of leftinternal carotid artery (arrow)

Letters


by lying to you about the treatment given toyou and claiming that it has some other pur-pose. For example, if a doctor thinks it wouldbe deleterious for you to have any morechildren he or she may give you an injectionof a long acting contraceptive while tellingyou that it is an immunisation.

“If you wish to guard against receivingtreatment without your knowledge then youshould refrain from eating or drinkinganything offered to you by health profes-sionals unless you have closely supervised itspreparation, or unless it has first beensampled by somebody else, preferably amember of staff. You should also checkclosely any drug that you are offered—forexample, by reading an account of its effectsin the British National Formulary. Thispublication also provides a description ofthe colour and shape of the tablets so thatyou can make sure that you are not beinggiven something other than what you havebeen told.

“The safest course of action if you wishto avoid being treated surreptitiously is tomaintain a constant state of vigilance and toregard all activities of medical and nursingstaff with the utmost suspicion. Try to haveas little to do with them as possible, and ifyou are at all concerned that somethingunderhand may be going on then do nothesitate to make good your escape.”

That should put everybody in thepicture.David Curtis Consultant psychiatristRoyal London Hospital, London E1 1BB

1 Kellett JM, Griffith D, Bell A, Short J, Adshead G. A nurseis suspended. BMJ 1996;313:1249-51. (16 November.)

2 Suspension of nurse who gave drug on consultant’sinstructions [letters]. BMJ 1997;314:299-301. (25 January.)

Allowing relatives to witnessresuscitationEditor—In response to the ResuscitationCouncil’s proposals,1 Kevin Stewart andLeslie Bowker cautioned doctors againstinviting relatives of adult patients to witnessresuscitation.2 Their arguments are that therelatives have no legal rights and that a com-plaint may arise from a breach of patientconfidentiality. The risks they mention,although correct, are far outweighed by thepotential benefit to the relative or closefriend.

The Resuscitation Council’s recommen-dations presuppose that legal principles arecontingent and not absolute and areoutweighed by humane considerations. Forsome people it is extremely distressing to beseparated from their loved one at this time,3

and there is also some evidence that beingthere, even briefly, can be beneficial togrieving.4 If aware, the patient may also ben-efit. I know of no evidence that most patientswould not want their relatives present,assuming the relative wants to be there.

Obtaining the patient’s consent is obvi-ously impossible, but I feel it would be wrongto deny all relatives the potential benefit of

being with their loved one at the end forwhat is a theoretical risk of complaint. How-ever, any known pre-existing wishes of thepatient should be respected. The Resuscita-tion Council’s guidance provides safeguards,including that the patient takes priority.1

In emergencies we act in good faith to dowhat we believe is best for patients and theirrelatives.

In the current state of knowledge,the Resuscitation Council project team feltthat the legal and confidentiality issuesare generally outweighed by humaneconsiderations.Chris A J McLauchlan, Chairman of ResuscitationCouncil project teamAccident and Emergency Department, RoyalDevon and Exeter Hospital (Wonford), ExeterEX2 5DW

1 Wise J. Push to allow relatives to witness resuscitation.BMJ 1996;313:899. (12 October.)

2 Stewart K, Bowker L. Resuscitation witnessed by relatives.BMJ 1997;144-5. (11 January.)

3 Adams S, Whitlock M, Bloomfield P, Baskett PJF. Shouldrelatives watch resuscitation? BMJ 1994;308:1687-9.

4 Gregory CM. I should have been with Lisa as she died.Accident and Emergency Nursing 1991;3:136-8.

5 Doyle CJ. Family participation during resuscitation: anoption. Ann Emergency Med 1987;16(6):107-9.

Medicine, postmodernism, andthe end of certainty

Postmodern philosophy offers a moreappropriate system for medicine

Editor—Paul Hodgkin equates postmod-ernism with the death of belief.1 A moreappropriate interpretation is that post-modernism argues against universal belief.1

Interpreting postmodernism according tothe statement “where one version of truth isas good as another anything goes” is in facta mistake. Postmodernism is an attack ondogma. It is for openmindedness and aboutother ideas and views. Postmodern philo-sophy would argue that there can be a truthwhich is not based on any particular beliefsystem but on an agreed basis within asociety at a particular time.2 As Hodgkinputs it, “To the postmodern eye truth isnot out there waiting to be revealed butis something which is constructed bypeople.”

I suspect that Karl Popper would arguestrongly against evidence based medicine asa sole way forward. After all, the evidence isbased only on our current value systems,which can dramatically alter with newadvances in our understanding of nature.3

The postmodern philosophy in fact offers usa more appropriate system for medicine.The truth is based on a system agreed bysociety in terms of current scientific knowl-edge, cost, and the personal preference ofindividual patients. We have to question whyso many of our patients resort to so calledalternative therapy with minimal or noscientific evidence. It is because we try toimpose our truth on our patients, who mayhave their own truth. We have to becomemore postmodernist if we are going to treat

our patients appropriately rather thansimply impose our dubious value systems onthem.Nick Raithatha General practitionerHealth Centre, University of East Anglia, NorwichNR4 7TJ

1 Hodgkin P. Medicine, postmodernism, and the end ofcertainty. BMJ 1996;313:1568-9. (21-28 December.)

2 Cahoone I. From modernism to post modernism. Oxford:Blackwell, 1996.

3 Popper K. Conjectures and refutations. London: Routledge,1963.

Doctors have a duty to remain truepatient advocates

Editor—Paul Hodgkin is right to sound thealarm.1 If indeed “one version of the truth isas good as another,” then anything goes.1

The postmodern world of fuzziness andfutility which he paints rejects the possibilityof absolutes, whether of scientific truths orguaranteed treatment outcomes. This denialof overarching frameworks, expressed byJean-Francois Lyotard as a “suspicion ofmeta-narratives,” inevitably leads to thedemise of traditional orthodoxies, whetherphilosophical or medical.2

Initially this view seems quite attractive:free agents uncluttered by old hierarchiesand anachronistic practices. As certaintiespass away, competing ideologies and treat-ments step into the health supermarket. Inthis brave new world who needs licences andmandates to practise? Free market competi-tion dictates that “licensure should be elimi-nated as a requirement for the practice ofmedicine.”3 All providers are welcome to bid,and the consumer decides.

Such deregulation shatters the hierar-chical division between scientific and alter-native medicine. The traditional autonomyof the medical profession collapses ascorporations and competing therapistsinvade the newly created health market.Indeed, this is already beginning to happen.4

Here the consumer picks and mixes medicalcare options from the shelves of enticingtherapies. This is where McDonalds meetsMrs Thatcher.

The beauty of traditional general prac-tice has been its ability to be both patientadvocate and health service gatekeeper. Thispotentially compromised dual role hassurvived intact to date. Such a possibilityreflects a meta-narrative that speaks oflicensed doctors who act with integrity, seek-ing to balance patient need, service capabil-ity, and scientific evidence. Such doctorscontinue their privileged role by consensusand through agreements over what is trueand trustworthy.

Although the postmodern world willpersist in challenging all our assumptionsand practices,5 we have a duty to remain truepatient advocates, not least as robustdefenders of the scientific medical heritage.We must firmly resist the idea that nothingcan be known, that truth is what you make it.That way leads to superstition and a newdark age.James Harrison General practitionerChaveley Park Medical Centre, Belmont, DurhamDH1 2UW

Letters



2 Turner BS. The interdisciplinary curriculum: from socialmedicine to postmodernism. Sociology of Health and Illness1990;12:1-23.

3 Friedman M. Capitalism and freedom. Chicago: Universityof Chicago Press, 1962.

4 Secretary of State for Health. Primary care: the future.Choice and opportunity. London: HMSO,1996.

5 Harrison J. Is this what we really want? BMJ 1996;313:1643. (21-28 December.)

Studies of environmental risk must notbe subject to bias from pre-existingbeliefs

Editor—The editorial by Paul Hodgkin onthe impact of postmodernism on medicineraises issues that are particularly relevant tothe practice of public health medicine.1 Howcan public health doctors hope to persuadepeople of the dangers of some forms ofbehaviour or environmental exposure andreassure those same people about the com-parative harmlessness of other behavioursand exposures when certainty is dead?

For every viewpoint the media canusually identify someone or some groupwith the opposite view. Postmodernistswould argue that both viewpoints areequally valid, whatever the evidential basis ofthose views. This is manifest most starklywhen the beliefs of professional and laycommunities differ, as in the assessment ofhealth risks associated with environmentalpollution incidents.2 3 These differenceswould merely be an indication of the needfor improved public education were it notthat prior lay knowledge or belief poses thereal risk of undermining the basis of profes-sional public health knowledge. Any epide-miological study that is conducted toinvestigate the impact of an environmentalinfluence on human health could be subjectto bias if the study population had alreadydecided the issue. Thus in a postmodernistworld epidemiological studies may serveonly to reinforce prevalent viewpointsrather than to determine absolute truths.This was most clearly shown when a strongassociation between tap water consumptionand adverse pregnancy outcomes was iden-tified in California.4 Eventually the authorssuggested that the likely explanation for thisassociation was that intense local anxietystimulated by press reports on unsafe waterhad biased reporting of exposure. The criti-cal issue for epidemiology in postmodernistsociety is how to design studies of environ-mental risk that are not subject to bias frompre-existing beliefs, either from lay orprofessional communities. But perhaps thiswas always the issue.Paul R Hunter Consultant in medical microbiologyand communicable disease controlPublic Health Laboratory, Countess of ChesterHealth Park, Chester CH2 1UL


2 Coggon D. Camelford revisited. BMJ 1991;303:1280-1.3 Brown P. Popular epidemiology and toxic waste contami-

nation: lay and professional ways of knowing. J Health SocBehav 1992;33:267-81.

4 Wrensch M, Swan SH, Lipscomb J, Epstein DM, NeutraRR, Fenster L. Spontaneous abortions and birth defectsrelated to tap and bottled water use, San Jose, California,1980-1985. Epidemiology 1992;3:99-103.

Complex systems result in a new kind offundamental uncertainty

Editor—Paul Hodgkin asks why medicinestill retains a modernist view of knowledgewhile many other subjects have abandonedthe search for an objective understanding ofexternal reality in exchange for therelativism and uncertainty ofpostmodernism.1 He suggests that this isbecause medical technologies have in factdelivered many universally applicableresults, particularly within the biomedicalsciences. In contrast, clinical practice hashad to confront an increasing number ofsocial, ethical, political, and economic issueswhose nature remains unclear. Instead ofscience removing uncertainty, the failure ofscience in these areas has increaseduncertainty by raising the possibility thatthere is no truth “out there” waiting to bediscovered. Postmodernism, with its discur-sive, qualitative methods, may simply movein to take over where quantitative methodshave failed.

In recent years there has been increas-ing interest in complex systems—that is,systems consisting of large numbers ofinterconnected non-linear units.2 Biologicaland psychosocial systems are all examplesof complex systems. An essential feature ofthese systems is that the properties of thewhole cannot be predicted from theproperties of the component units.Instead, properties emerge. Placed withinan evolutionary framework, complexadaptive systems evolve, selected on thebasis of their emergent properties. Theirstructure is not designed: it is self organis-ing. There are no internal mechanisms tobe discovered.

These observations have major implica-tions for research into clinical practice. Theyindicate that there may be a limit to empiri-cal science. As many aspects of clinical prac-tice are emergent properties of complexsystems, there is no “thing out there” to bediscovered. We are likely to be dealing with anew kind of fundamental uncertainty.

Interestingly, complexity could provide atheoretical underpinning to postmodern-ism. Work within the artificial life paradigmsuggests that the biological apparatus andcultural structures needed for the genera-tion and perception of higher level emer-gent behaviour—for example, language,meaning, or metaphor—could simply haveevolved without the need for any internalsymbolic representational schema.3 This isin keeping with postmodern ideas that allmeaning is constructed by a hermeneuticprocess and that all theories are metaphorswhich evolve as sociohistorical movements,selected by their rhetoric strength ratherthan any concept of proof.4

Complexity has been described as thethird epoch of science, after mathematicsand empirical science. The future of medicalscience may be uncertain.David P Frost Senior lecturer in psychiatryDepartment of Psychiatry, University CollegeLondon, Middlesex Hospital, London W1N 8AA


2 Coveney P, Highfield R, Frontiers of complexity—the searchfor order in a chaotic world. London: Faber and Faber, 1995.

3 Boden M, ed. Autonomy and artificiality. In: Philosophy ofartificial life. Oxford University Press, 1996.

4 Soyland AJ. Psychology as metaphor. London: Sage, 1994.

Rationing breast reductionsurgeryEditor—We are concerned with restrictedaccess policies for breast reduction by healthcommissions described by MargaretSomerville.1 Certain health commissionsrequest referral to other specialties beforefunding a plastic surgery consultation. Westudied the rationale of patient selection bythis method by examining the symptoms ofpatients whose outcome was determined bythe health commission.

Between February 1995 and January1996, 58 patients from three healthcommissions with restricted access policieswere referred by their general practitionersto Salisbury District Hospital for breastreduction. A standard letter was returned tothe general practitioners explaining that thereferral could not be accepted withoutapproval of the health commission. In 22cases the general practitioners took no fur-ther action; in the remaining 36 cases thehealth commission was contacted. In 16cases the health commission accepted thegeneral practitioner’s opinion and fundedthe procedure. In three of the 36 cases thehealth commission refused outright to fundthe request. In these patients the symptomsreported in the referral letter were notdifferent to those of the other patients. In 16of the 36 cases the health commissionrequired a referral other than for plasticsurgery (15 psychiatric opinions and 1orthopaedic opinion). Symptoms ofpatients funded for a psychiatric referraldiffered significantly from those fundeddirectly for plastic surgery referral onlyon “ridicule” (P = 0.04, Fisher’s exact test),and not on pain, posture, breathing,intertrigo, asymmetry, clothes, sport, anxi-ety, depression, unhappiness/emotionality,or embarrassment.

Thirteen of 15 psychiatric opinions werefavourable, 1 was unfavourable, and 1unknown. The health commission accepted11 of the favourable reports but refused tofund two. The favourable orthopaedic opin-ion was accepted. Finally, 28 of the 58patients were funded for a plastic surgeryconsultation.

There seem to be discrepancies inrestricted access policies. The grounds foroutright rejection of patients by the healthcommission were not specified. The groundsfor referral to another specialist wereunclear as symptoms of patients funded withgeneral practitioner’s support were the sameas those requiring psychiatric referral. Since14 of 15 referrals to other specialtiesresulted in favourable reports, psychiatricand orthopaedic resources are wasted inproviding consultations which almost invari-ably led to consultations with the plastic

Letters


surgeon. Direct referral from a generalpractitioner to plastic surgeon seems to pro-vide the most satisfactory and cost effectiveapproach. Psychological assessment, ifrequired, could follow and be performed bya clinical psychologist or psychiatrist experi-enced in assessing this patient group. Breastreduction surgery has been shown to bebeneficial to patients,2 3 and a uniform policyshould be established to ensure equitableaccess.N Horlock Research fellow in plastic surgeryRaft Institute for Plastic Surgery, Mount VernonHospital, Northwood, Middlesex HA6 2JRR P Cole Consultant plastic surgeonL F A R Rossi Consultant plastic surgeonOdstock Centre for Burns, Plastic and MaxillofacialSurgery, Salisbury District Hospital, Salisbury,Wiltshire SP2 8BJ

1 Somerville M, Radford G, Hewis N. Should breast reduc-tion surgery be rationed? Interventions requested forpsychological reasons should be studied. BMJ 1986;313:1478-9. (7 December.)

2 Cole RP, Shakespeare V, Shakespeare P, Hobby JA.Measuring outcome in low-priority plastic surgerypatients using quality of life indices. Br J Plast Surg1994;47:117-21.

3 Klassen A FR, Jenkinson C, Goodacre T. Should breastreduction surgery be rationed? A comparison of thehealthcare status before and after treatment: postal ques-tionnaire survey. BMJ 1996;313:454-7.

AuthorshipSee editorial by

Smith and p 1009

Guidelines exist on ownership of dataand authorship in multicentrecollaborations

Editor—Richard Smith and Richard Hortonhave expressed concern regarding researchmisconduct and, in particular, misappropria-tion of information and ideas and definitionsof authorship.1 2 Issues of data ownership andauthorship are particularly problematic formulticentre projects, where the need to “pub-lish or perish” among the increased numberof researchers is compounded by the require-ment for research accountability. Multicentrecollaborations are becoming increasinglycommon: they give greater breadth ofexperience and expertise in developing andimplementing protocols, facilitate recruit-ment of subjects, and allow a greater varietyof setting to enhance generalisability orresults. Despite its importance, little has beenwritten about this area.3-5

The National Psychosis ResearchFramework (currently coordinated atPRiSM, Institute of Psychiatry) recentlyestablished guidelines for multicentrecollaboration.

(1) A publication plan should be con-structed early on to avoid overlap ofcontents and ensure fair allocation ofauthorship. Prior written agreement must beobtained from the principal investigators ofall centres whose data is to be used in anypaper, and policy on authorship, corre-spondent, and acknowledgement for eachpaper should likewise be agreed.

(2) Individuals interested in data col-lected at other sites should circulate theirpublication proposals to those sites, giving areasonable specified time period in which torespond.

(3) Authorship should reflect relativecontributions to the writing of the paper aswell as to the analysis, design, and conduct ofthe study. It may be appropriate for somecontributions to be recognised by anacknowledgement.

(4) The final draft of any paper must becirculated to all authors, whose writtenagreement regarding content and targetjournal must be obtained before eachsubmission for publication. Consent mustalso be obtained from those whose contribu-tion has been acknowledged.

(5) When a public presentation of datais to be made and data are taken from apaper already accepted for publication,collaborating authors must be acknowl-edged. Where the data for presentation aretaken from a paper still in preparation, per-mission must be obtained from the relevantauthors for their findings to be presentedand the authors likewise acknowledged.Agreement may be given in advance for cer-tain categories of presentation.

(6) All papers which combine data frommore than one site should state theaffiliation to the wider research collabora-tion. Papers should be distributed to allmembers.Andrew Barker Senior research fellow in old agepsychiatryUniversity of Southampton, Thornhill ResearchUnit, Moorgreen Hospital, Southampton SO30 3JBRichard A Powell ResearcherUniversity of Exeter, Department of Mental Health,Wonford House Hospital, Exeter EX2 5AF

1 Smith R. Time to face up to research misconduct. BMJ1996;312:789-90.

2 Horton R. Time to redefine authorship. BMJ1996;312:723.

3 Hogg RJ. Trials and tribulations of multi-centre studies:lessons learned from the experiences of the SouthwestPediatric Nephrology Study Group. Pediatr Nephrol1991;5:348-51.

4 Kassirer JP, Angell M. On authorship and acknowledge-ments. N Engl J Med 1991;325:1510-2.

5 Hanson SMH. Collaborative research and authorshipcredit:beginning guidelines. Nurs Res 1988;37:49-52.

Team approach to assigning authorshiporder is recommended

Editor—Deciding the order of authors onresearch papers is a recognised problem.1-5

Currently, authorship order cannot be inter-preted by readers and editors. The last posi-tion often carries more status.3 In somepapers the senior investigator is named last,in others it is the head of the laboratory ordepartment, and in others it is the personwho contributed least.

Burman recommended that theresearch supervisor should be secondauthor and the head of laboratory last,4

while Riesenberg and Lundberg proposedthat order should reflect contribution.1 Mostwriters agree that the person most closelyassociated with the research should be thefirst author.1-4 The Council for BiologyEditors recommend that each journal speci-fies criteria for authorship order, butjournals tend to leave this to authors. To ourknowledge no practical scheme for allowinggroup consensus on authorship order hasbeen published, although there are tech-niques for calculating credit.

We undertook a piece of participativeresearch to assign authorship order to ourwork5 by means of peer judgments. We com-piled a list of contributions involved incarrying out our research (see box). Teammembers spoke briefly about their contribu-tions and ticked the list appropriately. Usingthis list and background knowledge aboutthe development and execution of theproject, each team member compiled threerankings of colleagues’ contributions,excluding themselves. The first was for theproject as a whole, and the others for each oftwo papers. A top rank was given eightpoints, a second seven, and so on. Wecollated the rankings to give a total score foreach team member, and all nine authors dis-cussed and agreed the resulting order.Despite initial trepidation and discomfort,we agreed that the outcome of this processwas fair.

Authors’ possible contributions toresearch project• Impetus or initiative• Preparation• Planning meetings• Sampling• Questionnaire design• Administering the survey• Interviewing subjects• Coding questionnaire• Qualitative analysis• Quantitative analysis• Drafting paper• Commenting on drafts• Attending meetings• Conference preparation

Our approach makes each author’sinvolvement explicit within the team anduses an anonymous democratic process. Asthe rankings can be secret, abuse of power,status, and reputation are minimised. Thismay help to reduce difficulties experiencedby researchers. Determining authorship inadvance may reward those who promise atthe expense of those who deliver, whereasour approach reflects the team’s view ofactual contributions. We suggest that otherstest this process for themselves.Authorship order reflects peer group views on over-all academic and practical contribution to theproject and is therefore rank order.R S Bhopal Professor of epidemiology and public healthJ M Rankin Senior research associateDepartment of Epidemiology and Public HealthE McColl Senior research associateCentre for Health Services ResearchR Stacy Lecturer in medical sociologyP H Pearson Lecturer in primary care nursingE F S Kaner Research associateDepartment of Primary Health CareL H Thomas Research associateCentre for Health Services ResearchB G Vernon Lecturer in the ethics of health careDepartment of Primary Health CareH Rodgers Senior lecturer in stroke medicineDepartments of Medicine and of Epidemiology,School of Health Sciences, University of Newcastle,Newcastle upon Tyne NE2 4HH.

Letters


1 Riesenberg D, Lundberg G. The order of authorship:who’s on first. JAMA 1990;264:1857.

2 Day RA. How to write and publish a scientific paper. 4th ed.Phoenix, AZ: Oryx, 1994.

3 Burman KD. Hanging from the masthead. Ann Intern Med1982;97:602-5.

4 Huth EJ. How to write and publish papers in the medicalsciences. 2nd ed. Baltimore: Williams and Wilkins, 1990.

5 Bhopal RS, Rankin M, McColl E, Thomas LH, Kaner EFS,Stacy R, et al. The vexed question of authorship: views ofresearchers in a British medical faculty. BMJ 1997;314:1009-12. (5 April.)

New bill offers control formentally disordered offendersEditor—In his recent editorial DerekChiswick commented that clause 36 of theCrime (Sentences) Bill “runs contrary tomodern concepts of psychiatric practice.”1

Central to this debate is the issue of respon-sibility, the extent to which a psychiatricpatient can be held responsible for a crimi-nal act. Policy in this area has followed therecommendations, in 1975, of the Butlercommittee: “Prosecution should be seen asa last resort,”2 which were subsequentlyreaffirmed by the Read review in 1992:“Wherever possible receive care and treat-ment from Health and Social Servicesrather than in the Criminal Justice System.”3

The bill is a major change from thisposition; it offers the court the opportunityto combine treatment and punishment andthus to consider the issue of a patient’sindividual responsibility for an offence.

The change should be welcomed. Themajority of high profile failures of commu-nity care have involved patients with ahistory of offending behaviour. They rarelyface prosecution because of the currentpolicy. I have looked after several patientswho have assaulted members of the public,staff, and other patients, both in the commu-nity and in hospital. They have rarely, if ever,been prosecuted. This problem hasincreased in recent years with the develop-ment of community care and the wide-spread availability of psychoactive drugs.Community care offers new opportunitiesand freedoms for patients, but with it comesa responsibility for patients to cooperate andhelp in the management of their own illness.Current practice often seems to placepsychiatric patients above the law, withoutrisk of prosecution for a range of offences. Inthis context prosecution may have benefitsbeyond the immediate protection of thepublic. It offers the prospect of encouragingpatients to comply with their treatment regi-men, whether this is compliance withprescribed medication or avoidance ofpsychoactive drugs that might themselvestrigger a relapse; it may also have some valuein increasing patients’ insight into theseriousness of their behaviour when ill, andit emphasises the responsibility of thepatient to assist in cooperating with theirown care and treatment.

The opportunity should be seized toprovide a legal framework that will underpinand strengthen the controls for managingmentally abnormal offenders in the commu-nity. This would include wider prosecution

of offences, greater use of sentences such as“probation with treatment as a condition,”and a new section of the Mental Health Act1983, such as a community treatment orderwhich would build on the acknowledgedsuccess of restriction orders.4

D J Thompson Consultant psychiatristMalham House Day Hospital, Leeds LS2 9LN

1 Chiswick D. Sentencing mentally disordered offenders.BMJ 1996; 313:1497-8. (14 December.)

2 Home Office, Department of Health and Social Security.Report of the committee on mentally disordered offenders.London: HMSO,1975. (Butler report; Cmnd 6244.)

3 Department of Health. Home Office Review of health andsocial services for mentally disordered offenders and othersrequiring similar services. London: HMSO, 1992. (Reedreport; Cmnd 2088.)

4 Hedderman C. The supervision of restricted patients in thecommunity. London: Home Office Research and StatisticsDepartment, 1995. (Research findings No 19.)

Traders’ views on sale ofcigarettes to childrenEditor—If the high prevalence of smokingamong children1 is to be reduced then acomprehensive approach is required.2 Partof this approach is a reduction in theavailability of cigarettes. The Children andYoung Persons’ (Protection from Tobacco)Act 1991 strengthened the law on theprevention of sales of cigarettes to childrenaged under 16. We investigated traders’knowledge of and opinions on the healthissues related to smoking by children andthe legislation itself.3

We invited the 54 trading premises thatsold tobacco in three areas in northeastWales to participate in the study. Thirty twotraders agreed to do so and were visited by aresearcher and given a self completionquestionnaire.

Thirty respondents said that health wasthe main reason for the act, but threethought that smoking had no effect on chil-dren’s health and 10 did not know thatbecoming a smoker before the age of 15carries a higher risk of lung cancer thandoes starting smoking at the age of 25.4

Eighteen respondents thought that they hadsome responsibility for children’s health,and over half thought that they had a healthpromotion role for children: 21 by prevent-ing sales of cigarettes to them and four byinforming them of the dangers of smoking.Thirty one respondents said they would notpersonally sell cigarettes to children, but a

few mentioned exceptional circumstanceswhen they would—namely, if they were toldthat the cigarettes were for an adult, if theyreceived written parental consent, and if thecigarettes were for a parent in a family thatthey knew or for a disabled adult in a caroutside. The dilemma arising when youngpeople look older than they are wasmentioned. All the respondents were famil-iar with the act, but 13 did not know the cur-rent maximum fine (£2500 ($4000)).

Twenty nine respondents said thatchildren attempted to buy cigarettes fromtheir premises. Seven showed these childrenthe statutory notice relating to sales of ciga-rettes to children under 16, and 11 asked foridentification as evidence of age. Mostpremises had guidelines for sales staff, butowners generally showed more awarenessand commitment than managers. Manytraders thought that the government couldhelp them in various ways (table 1).

As Devine and Vickers suggested,traders would benefit from education aboutthe reasons for the age limit for sales oftobacco as well as about the act itself.5 Ourrespondents also believed that they wouldbe helped by action by the government,especially the introduction of identity cardswith photographs.

We thank the trading standards department andall the participants for their help in this study. AC’swork is supported by the Cancer ResearchCampaign (grant No CE1055) and DJ’s work wassupported by the Clwydian Community Care NHSTrust.Delwyth Wyn Jones Programme manager, preventionof smokingHealth Promotion Service, Mold, FlintshireCH7 1XBAnne Charlton Professor of cancer health educationCancer Research Campaign Education and ChildStudies Research Group, School of Epidemiologyand Health Sciences, University of Manchester,Manchester M13 9PT

1 Diamond A, Goddard E. Smoking among secondary schoolchildren in 1994. London: HMSO, 1995.

2 Royal College of Physicians. Smoking and the young.London: RCP, 1992.

3 Edwards D. Sales of cigarettes to children in Clwyd.Manchester: University of Manchester, 1995. (MSc thesis.)

4 US Department of Health and Human Sciences. Thehealth consequences of smoking: cancer. A report by the surgeongeneral. Rockville, MD: Office of Smoking and Health,Public Health Service, 1982.

5 Devine MJ, Vickers J. Shopkeepers should be educatedabout smoking among children. BMJ 1996;312:441.

Table 1 Changes in law relating to sales of cigarettes to children that traders would like to seegovernment implement. Figures are numbers (percentages)

Newsagents(n=14)

Grocers(n=9)

Petrol stations(n=6)

Offlicences

(n=3)

Totalsample*(n=32)

Issue identity cards with photographs 12 (86) 9 (100) 6 (100) 3 (100) 30 (94)

Change current law (make it illegal for children tobuy cigarettes, or raise age limit)

6 (43) 6 (67) 2 (33) 2 (67) 16 (50)

Ban smoking in public places 5 (36) 5 (56) 2 (33) 0 12 (38)

Ban cigarette advertising 2 (14) 3 (33) 2 (33) 0 7 (22)

Increase tobacco prices 0 3 (33) 2 (33) 0 5 (16)

Other 3 (7) 0 0 0 3 (9)

*Some respondents gave more than one answer.

Letters


Head up tilt testing has a placein distinguishing certainconditions from epilepsyEditor—In their editorial J W A S Sanderand M F O’Donoghue place great relianceon clinical characteristics to differentiatebetween epilepsy, convulsive syncope,and non-epileptic attacks.1 This seemsunwise because the diagnostic success rateeven of specialists may be as low as 50% onthe basis of history and examination alone.2

Furthermore, direct observation of seizuresby trained staff in a specialised inpatientunit resulted in only 80% diagnosticaccuracy.2

One important omission in the editorialwas the failure to mention head up tilttesting, which has assumed an importantrole in diagnosing syncope in recent years.3

Even when used in reported cases ofepilepsy head up tilt testing has been shownto be valuable in detecting convulsivesyncope and seizures of psychogenic origin.3

We recently reported four cases of primarycardiovascular disease misdiagnosed asepilepsy.4 One of these patients was a middleaged woman with classic tonic-clonic sei-zures who had 38 seconds of asystole onhead up tilt testing, followed by a short livedgeneralised seizure. Her seizures resolvedwith permanent pacing. We subsequentlyidentified two additional patients withsuspected epilepsy who had profound vaso-depressor reactions to head up tilt testingassociated with reported seizures. Anotherpatient with so called intractable epilepsydeveloped violent, semipurposeful shakingof the limbs within two minutes of head uptilt testing but without associated haemody-namic or electroencephalographic changesand was concluded to have non-epilepticattack disorder.

In the first three patients the seizureactivity was clearly genuine, but the causewas mistaken. The cause was global cerebralhypoxia, which in turn was the result of glo-bal cerebral hypoperfusion due to suddenprofound hypotension. What must berecognised is that cerebral hypoperfusionmay cause what look like epileptiformmovements. To both lay and skilled observ-ers this phenomenon may be indistinguish-able from epilepsy caused by a generaliseddischarge of cerebral neurones while cer-ebral perfusion is preserved. If seizure activ-ity due to hypotension is mistaken forepilepsy many young patients who areprone to reflex syncope may be labelled epi-leptic. This has profound effects on pros-pects for life, work, marriage, childbearing,and driving. Indeed, data on patients with afirst seizure show that most do not haveepilepsy confirmed by electroencephalo-graphic abnormalities.5 Many of thesepatients might have reflex syncope with sei-zure activity that is misinterpreted as aprimary cerebral disturbance. Althoughreflex syncope is not always benign, its diag-nosis has far less severe connotations than adiagnosis of epilepsy.

It is important that reflex syncope isconsidered in patients presenting withseizure activity. Although far from perfect insensitivity and specificity, head up tilt testingprovides a tool for objective assessment ofsuch patients and avoids relying solely onthe well documented limitations of a clinicalevaluation.Amir Zaidi Research fellowLawrence Cotter Consultant cardiologistAdam Fitzpatrick Consultant cardiologistAutonomic Function Department, ManchesterHeart Centre, Royal Infirmary, ManchesterM13 9WL

1 Sander JWAS, O’Donoghue MF. Epilepsy: getting thediagnosis right. BMJ 1997;314:158-9. (18 January.)

2 King DW, Gallagher BB, Murvin AJ, Smith DB, MarchsDJ, Hartlage LC, et al. Pseudoseizures: diagnostic evalua-tion. Neurology 1982;32:18-23.

3 Grubb BP, Gerard G, Roush K, Temesy-Armos P, Elliott L,Hahn H, et al. Differentiation of convulsive syncope andepilepsy with head-up tilt testing. Ann Intern Med1991;115:871-6.

4 Zaidi A, Cotter L, Fitzpatrick A. Sudden unexplaineddeath. Lancet 1997;349:135.

5 Hopkins A, Garman A, Clarke C. The first seizure in adultlife. Lancet 1988;i:721-6.

Prognosis for babies of lessthan 28 weeks’ gestationEditor—Win Tin and colleagues providefurther help for parents at risk of deliveringbabies of less than 28 weeks’ gestation.1

They concluded that “these data should notbe used to establish an arbitrary ‘threshold’below which resuscitation should never beoffered at birth” but did not mention howmany of the 342 babies born at 22 to24 weeks were resuscitated and admittedto intensive care and how many were justkept comfortable and cuddled by theparents. Although data on profound andsevere handicaps were useful, it would havebeen more important to know aboutthe incidence of normal outcome at eachgestation age.

In clinical practice a major problem forexpectant parents is not being supplied withconsistent and realistic information andadvice by obstetricians, midwives, andneonatologists. The data of Tin andcolleagues support the need for parents toknow that babies born at 26 weeks and ear-lier are at very high risk of death orhandicap.2 Communication, as in a recentcase of a baby of 23 weeks’ gestation,3 isespecially difficult in the emotionallycharged atmosphere of delivery suites,when parents can have problems under-standing and remembering, no matter howwell the information had been given. Overthe past two and a half years I haveroutinely recorded (with consent) myconversations with parents using a “walk-man” tape recorder, a tie-clip microphone,and standard C60 cassettes. At the end ofthe conversation I give the tape recorderwith the cassette to the parents so that theycan listen to the information at leisure (andrepeatedly if necessary). Parents seem tobenefit from such a simple adjunct in com-munication. There were two parents who on

listening to the recording of the conversa-tion claimed that the conversation nevertook place.T H H G Koh Senior specialistCanberra Clinical School, Canberra Hospital,Canberra 2605, Australia

1 Tin W, Wariyar U, Hey E. Changing prognosis for babiesof less than 28 weeks’ gestation in the north of Englandbetween 1993 and 1994. BMJ 1997;314:107-11.(11 January.)

2 Koh THHG. Counselling parents of extremely prema-ture babies. Lancet 1997;349:289.

3 Kmietowicz Z. Premature baby was not put on ventilator.BMJ 1996;313:963. (19 October.)

Teleconferencing should beextended to include jobinterviewsEditor—Twice in the past five years I havemade trips from Australia to Britain for thesole purpose of attending an interview. Onboth occasions I was unsuccessful. Thisprompts me to suggest that professor D M BHall’s indications for teleconferencing1

should be extended to include job inter-views. He notes that in the British PaediatricAssociation’s experience teleconferencinghas reduced the overall cost of holding ameeting by half to two thirds in view of thereduction in “travel and catering costs.” Thesavings available by holding teleconferenceinterviews for overseas candidates are fargreater. For example, my latest trip entailedtwo weeks’ absence from work (half myannual leave), 10 days’ accommodation andsubsistence in the United Kingdom, in addi-tion to the cost of a transglobal flightarranged at short notice.

According to Whitley Council stand-ards, NHS hospitals have been liable fortravel costs for interviewees from their pointof entry to Britain, and candidates have hadto accept the cost of their journey to Britainas the price of gaining overseas experience.Teleconferencing makes many of these costsunnecessary. One would hope that anyreasonable prospective employer would notinsist on the candidate making such a finan-cial commitment when a viable alternativeto face to face interview exists.

Some employers may think that this isan issue for the candidate and his or herpurse. It may, however, become more of anissue for the interviewing hospital. TheWhitley Council standards discriminateagainst all candidates from overseas, includ-ing those from other countries within theEuropean Union. If this breaches theEuropean agreement on free movement ofgoods and labour employers may find thatthey have a responsibility for reimbursingtravel expenses from any point within theEuropean Union. This must be an incentivefor NHS employers to utilise teleconferenc-ing for interviews.Graham Briars Paediatric gastroenterologistRoyal Children’s Hospital Brisbane, Herston,Queensland 4029, Australia

1 Hall DMB. Telephone conferencing. Arch Dis Child1996;75:460-1.

Letters


Documents

Authorship. Team approach to assigning authorship order is recommended