Tobacco Documents Online

Le~e~ potentially serious complication of a diag- nostic procedure.' A S~rachan Specialist regisrrm" J Tr~n Comuttam ana~thaist Directorate of Anaesthesia. Doncaster Royal Infmnary, Doncaster DN2 5LT I Broad/t'/SA, Ful/er NE. Lumbar ptmcmre needn't be a bendable. BM] 1997;515: IS24.5. (22 November.) 2 Canon D, Selma M. Choosing the beat needle for d/agnm~c lumbar ptmcmre. ~Vran~ [9~6;47'~5-7 3 HardmanjG, Gajraj NM. Epiduml blood patch. ~rfHo~p ,'t4~ 1996.,56..208-9. 4 Reid JA. Tho~ourn J. Headache after spinal arta~thesia. Br f Aaattth 1991,'67:674-7. Epidural blood patching can he used to /rent headache E0rroa--The editorial by Broadiey and Fuller cal/s for a shift to atraumatic needles in diagnostic lumbar puncture.t The advan- tages such needles confer in terms of a 'reduced incidence of headache after lumbar puncture are well known to anaesthetists and were discussed in a BMJ editorial four years ag~x= It is unfortunate that this message has still hot reached all concerned, which suggests that the complication has not been taken sufliciendy seriously. Although headache after lumbar punco turn tends to be self tim}sing, patients may have incapacitating pain in the interim. In a signiticant minority the headache may persist for more than six week~~ Most importantly, the low cerebrospinal fluid pressure d'mt underlies headache after lumbar puncture is sometimes associated with more severe consequences, including subdural hae- matoma and seizure~* 4 In 1993 Reynolds identified 81 references in Medline to we have identified a further 14 such reports to date (list available on request). These compli- catiom often occur when headache has been misdiagnnsed or ineffectually treated. The authors do not comment on the actual treatment. Epidural blood patching, which necessitates referral to an anaesthetist, is effective in about 90% of patients, and repeating the procedure in those showing a poor respome raises this figure to over 98%? Epidmal blood patching restores normal cerebrospinal finid pressure and promotes repair of the dural defect:; its judicious use may avert the catastrophic outcomes that have occurred when postdural puncture headache was neglected. Reynolds therefore suggested that all patients who have "large needle" headache--and possibly also those w.ho have "fine needle" headache--for more than a week should undergo epidtwal blood • patching." In the absence of controlled trials, others may prefer to reserve epidural blood patching for those headaches severe enough to make mobilisatlon difficult or impossible or those associated with abnorrnalides of the auditory, visual or cranial nerve. We are concerned that referrals for epidural blood patching after diagnostic lumbar puncture are rare but incommensu- rate with the known incidence of headache. Another problem is that some junior doctors performing diagnostic lumbar puncture simply lack awareness that head- ache is a possible complication. Broadiey and Fuller have correctly }den- tiffed the need for all spec/aldes to use auau- matic spinal needles. Overall management of the complication a/so needs to be improved. This should include giving dear advice to patients, especially those dis- charged he.me early after dural puncture, and introducing local protocols to establish both indications and referral channels for early epidural blood patching. Mark R Nel S,n/or tog/sitar ht ana~thai~ Magill Departraent of Anaesthetic~ Chelsea and We~tmimter Hospital London SW10 9NH Neville Robinson C0nsu/~ ana~sth¢~t Department of Anaesthe~i~ Northwick Park Hospital, Harrow HA1 3UJ I Bmadley SA, Full~ GN. Lumbar puncture needn't be a headar.he..SMJ 1997'~ 15:1324-5. (22 November.) 2 Reynoid~ F. Dun~ puncture and headadae. 1993;3(]6:874-6. 3 MacArthur C. Lewi~ M. Knc~ EG. Acddental dth'al ~] t 99s'~oe'.88s-5. ,t Shad-or rE, jhaveri He. Cunniagham seizures after post-dm~1 puncaL,'e ~,eadadae. 1995;85:255-60. 5 Peterman SB. Pmunydosraphy headache:, a revie~ Radio/a~ 1996',200:7e5-70. ~ editorial has generated a large respome, and we received 14 other letter& Like the letters from Muidoon etal, me were f~om anaesthethts; the other five were from specia~ts in dermatology, intensive care, radiology, and neurology.--EJ3rrom Alcohol consumption eliminates rathe~ than prevents infection with HA/or/ EDrrog--Brenner et al conclude that alcohol comumpdon may have a protective effect against infection with HelicobacUr ~lori.l H py/or/infection is commonly acquired during childhood and spontaneous elimination may occur, usually in young children and elderly people, but it may be acquired and eliminated spontaneously throughout aduithoocLt Drinking alcohol is usuaJly an adult habit, and if alcohol does have a protective effect against infection with Hpy/or/we can assume that it eradicates the infection rather than prevents it. Alcohol comumpdon may therefore have a role only in those few people in whom spon- taneous elimination of the bacterium is thought to have ocxam'ed.~ If alcohol does protect against H py/or/ infection then the number of years for which it has been consumed is probably important, and not just the amount consumed, particu- Larly when overall consumption is low. Bren- ner et al do not, however, give data on lifetime alcohol consumption in their series. Finally, socioeconomic factors in chi/dhood, such as low social class, manual occupation of the parents, and measures of crowding have been consistendy identified as major risk factors for infection with H py/or/,~ but they were not adjusted for in the multivari- ate analysis in the authors' study. We recendy conducted a seroepidemio- logical study of Hpy/or/infection in 705 resi- dents in the rural town of Cire, a wine producing centre in southern Italy? A struc- tured questionnaire was administered to all subjectx The table shows data on alcohol consumption and its reladon to H Filer/ seropositivity. The seroprevalence of infec- tion was 630/0 (446/705) ove~all and was slighdy higher in wine drinkers (70% (58/83) than in non-drinkers (62% (~88 of 622)). By contrast, Hpy/or/seroposidvity was reduced in subjects who reported a lifetime alcohol consumption higher than 456 kg ethanol (64%, (27/42) v 76% (~ 1 of 41)). The differences were not sigrd~cant in a univari- ate or a multivariate analysi~ Serological testing for Hpylori has been validated, with tests indicating active infec- tion being used as a reference? Since spon- taneous elimination of the bacterium is uncommon,-" a positive result of a serologi- cal test for Hpy/or/should be considered as evidence of active infection when no specific ant}microbial treatment has been given. We believe that wine consumption is not a pro- tective factor against H pylor~ Different demographic and clinical wariables may account for discrepancies in findings. Frmacesco Lu,~a Sen/or res~rc.htr Maria Imeneo R~earch assistant Maria Maletta P~smreh assistant Francesco Pallone Pro#seer ofgaaromterolog~ Dipardmento di Medic}ha Sperimentale e Clinica. Universita di Catanzaro, Catanzaro, Italy 1 Plren~er I-L Rotherd~tcher D, Bode (3, Adler (3. Relation of smok/ng and a/£ohol and coffee consump~on to andre Hell.batter pylori infectiom cross sertional study. 1997;M5:1489-92. (6 Detember.) 2 Xia ~ Talley NJ. Natmal a~tuisldon and spontaneous dim}nation of Hdicobacter pylori infe~on: clinical imp}i- cat}ore. AmJ Gasmx.mm~ 1997;99:1780-7. 3 Webb PM, I~aight T, Greaves S, WLlsoa A. Newetl DG. Elder J. et aJ. Relad.on bets~.~m in£ection with Helicobacter and living conditions in chiJdhood: evidence for person l~rson wammmion in early life. BM] 1994"~08:750-3. 4 Luzza F, lmtmeo M, Maletta M, Palucdo G. Giancotti A. ~ne K et aL Seroep/derniology of Hdicobacter l~iori infection and hepatitis A in a rural area: evidence against a common mode of transmission. Gut 1997;41:164-8. 5 Cutler A. Havstad S, Ma ~ Blas~ M, Per~-l~r~z G, Schu- berx Z Accura~ of invasive and aoninvasive tests to diag- nose Helicobacter pylori infection. Relation of lifetime alcohol consumption to H py/ofi seropositivity Llfetlmo alcohol oonsumptlor~ No (%) eempolltlv~ (kg ethanol}" Be o# subbed'l= for H p¥1o~ Odds ratio (95% CI) Crtlde None 622 388 (62) 1.00 1.00 Range 46-1095 median 456 83 58 (70) 1,4 (0.85 to 2.3) 0.72 (0.38 to 1.35) • =;456 41 31 (76) 1.8 (0.86 to 3.76) 1.11 (0.48 to 2.53) >456 42 27 (64) 1.12 (0.57 to 2.I9) 0.49 (0.22 to 1.(]9) "Estimated from years of drinking and usual weekly intake of wine, assuming that 0.5 "titre of w=ne contains on average 50 g ethanol 1" Adiusted for age, sex, smoking, years of school education, occupation, dyspepsia, history of peptic ulcer, social crass, ch~ldhond crowding (people per room), and number ot s~hlings. BMI VOLL.'ME:tI6 L2$MARCH 1998 • 1019