340

infection, by haematogenous dissemination from dis-tant sites of suppuration, or from embolisation fromfoci of septic thrombophlebitis.26,27Most of the anaerobes we recovered were sus-

ceptible to penicillin G.28 A notable exception isB. fragilis, which was present in 15% of the cases.This organism is usually resistant to penicillin, butis sensitive to clindamycin and chloramphenicol.Drug selection in cases associated with an exclu-sively aerobic flora should ideally be based on in-vitro susceptibility tests. When aerobic and anaero-bic bacteria are present concurrently, combinationsof antimicrobial agents may be required, althoughthe need to treat all microbial isolates in these caseshas not been established. Drainage is a key facet oftherapy; in our experience open surgical drainagewith rib resection is usually required.

Requests for reprints should be addressed to J. G. B.,Veterans Administration Hospital, Sepulveda, California91343, U.S.A.

REFERENCES

1. Attebery, H. R., Finegold, S. M. Appl. Microbiol. 1969, 18, 558.2. Sutter, V. L., Attebery, H. R., Rosenblatt, J. E., Bricknell,

K. S., Finegold, S. M. Anaerobic Bacteriology Manual. De-partment of Continuing Education in Health Sciences, Univer-sity Extension and School of Medicine, U.C.L.A., Los An-geles, 1972.

3. Holdeman, L. V., Moore, W. E. C. Anaerobe Laboratory Manual.Virginia Polytechnic Institute and State University, Blacks-burg, Virginia, 1972.

4. Maes, U., Veal, J. R., McFetridge, E. M. J. thorac. Surg. 1935,i4, 615.

5. Penberthy, G. C., Benson, C. D. Ann. Surg. 1936, 104, 579.6 Nowak, S. J. G. Med. Clins N. Am. 1939, 23, 1355.7. Thompson, L. D., Edwards, J. C., Hoagland, C. L. Ann. intern

Med. 1940, 13, 1138.8. Ehler, A. A. Int. Abstr. Surg. 1941, 72, 17.9. Elfving, G. Acta chir. scand. 1954, 107, 454.

10. Finland, M., Jones, W. F., Barnes, M. W. J. Am. med. Ass.1959, 170, 84.

11. Lutz, A., Grooten, O., Berger, M. A. Strasburg Med. 1963, 14,119.

12. Simmons, E. M., Sauer, P., Elkadi, A., MacKenzie, J. W.,Almond, C. H. J. thorac. cardiovasc. Surg. 1972, 64, 578.

13. Snider, G. L., Saleh, S. S. Dis. Chest, 1968, 54, 12.14. Sullivan, K. M., O’Toole, R. D., Fisher, R. H., Sullivan, K. N.

Archs intern. Med. 1973, 131, 521.15. Weese, W. C., Shindler, E. R., Smith, I. M., Rabinovich, S.

ibid. 1973, 131, 516.16. Vianna, N. J. J. Am. med. Ass. 1971, 215, 69.17. Yeh, R. J., Hall, D. P., Ellison, R. G. Am. Rev. Resp. Dis. 1963,

88, 785.18 Stokes, E. J. Lancet, 1958, i, 668.19. Beerens, H., Tahon-Castel, M. Infection humaines, à bactéries

anaérobies non-toxigènes; p. 99. Brussels, 1965.20. Bechamps, G. J., Lynn, H. B., Wenzl, J. R. Mayo Clin. Proc.

1970, 45, 43.21. Lindskog, G. E. New Engl. J. Med. 1956, 255, 320.22. Stiles, Q. R., Lindesmith, G. G., Tucker, B. L., Meyer, B. W.,

Jones, J. C. Ann. thorac. Surg. 1970, 10, 37.23. Cohn, L. H., Blaisdell, E. W. Archs Surg. 1970, 100, 376.24. Bartlett, J. G., Gorbach, S. L., Finegold, S. M. Am. J. Med.

(in the press).25. Guillemot, L., Hallé, J., Rist, E. Archs méd. exp. Anat. Path.

1904, 16, 571.26. Bartlett, J. G., Finegold, S. M. Medicine, 1972, 51, 413.27. Tillotson, J. R., Lerner, A. M. Ann. Intern. Med. 1968, 68, 308.28 Martin, W. J., Gardner, M., Washington, J. A., Jr. Antimicrobiol

Agents Chemother. 1972, 1, 148.

Reviews of Books

Anesthesia and Uncommon Diseases

Pathophysiologic and Clinical Correlations. Edited byJORDAN KATZ, M.D., professor and associate chairman,department of anesthesiology, University of WisconsinMedical School, Madison, Wisconsin; and LESLIE B.KADIS, M.D., assistant professor of anesthesiology, Stan-ford University School of Medicine, Palo Alto, Califor-nia. Philadelphia and London : Saunders. 1973. Pp. 543.8.90.

MosT of the 26 contributors to this compendium areanaesthetists, and it is therefore surprising that theactual management of anaesthesia in the rare diseasespresented here is often very sketchy indeed compared withthe unnecessarily comprehensive accounts of the diseasesthemselves. For example, one and a half pages on Ray-naud’s disease is followed by the terse statement that noparticular type of anaesthesia is preferable. The styleand content of the chapters vary: some achieve their aimwhereas others are patently deficient. This book does notseem to have been adequately conceived, which is regret-table in view of the importance of the subject. One ofthe few chapters which comes off well is that on the man-agement of anaesthesia in drug addicts.

Lectures on the Methodology of Clinical ResearchMAX HAMILTON, M.D., F.R.C.P., F.R.C.PSYCH., F.B.PS.S,Nuffield professor of psychiatry, University of Leeds.Edinburgh: Churchill Livingstone. 1974. Pp. 197..65.

A SECOND edition of Professor Hamilton’s book is wel-come. Thirteen years have elapsed since it was first pub-lished, and the amount of fresh information that has

subsequently emerged must be immense; but selection ofnew material has been judicious, and the book remainscompact, clear, and readable. The author concentrates onclinical research and pays special attention to therapeutic

trials; but there is something to be said for a broaderapproach. Consideration of research activities in certainother medical and biomedical fields would have producedexamples of more circumscribed problems, beset withfewer variables and better suited to more rigorously con-trolled investigation. These would illustrate research pro-cedures applied under something at least approachingideal conditions; and they would also indicate that suchconditions are often artificial, contrived, and rarely repro-duced in clinical medicine, the peculiar features of whichmust determine what kinds of research methodology are(or are not) appropriate. Sir Peter Medawar once re-

marked that "Biology is complex, messy and richly vari-ous", and the remark applies equally to clinical medicineNot, however, in a defeatist sense: the special featuresof human medicine do not preclude clinical research butmerely make it more difficult, and increase the need forexpert guidance in planning and interpretation of the kindoffered by Professor Hamilton.

New Editions

Noise and Man. 2nd ed. By William Burns. London: JohnMurray. 1973. Pp. 459. E5.

Laboratory Techniques in Rabies. 3rd ed. Edited by Martin M.Kaplan and Hilary Koprowski. Geneva: World HealthOrganisation. 1973. Pp. 368. E6;$14.40; Sw. fr. 48.

Hewer’s Textbook of Histology for Medical Students. 9th ed.Edited by S. Bradbury. London: William Heinemann MedicalBooks. 1973. Pp. 465. S3.50.

Current Surgical Diagnosis and Treatment 1973. By J. EngelbertDunphy and Lawrence W. Way. Los Altos, California: Lange.1973. Pp. 1108.$14.

Notes on Radiological Emergencies. 2nd ed. By George Ansell.Oxford: Blackwell. 1973. Sl.

Genetics in Medicine. 2nd ed. By James S. Thompson andMargaret W. Thompson. Philadelphia and London: Saunderi.1973. Pp. 400. S3.85.

341

THE LANCET

Brain Damage and Brain DeathTHE importance and the difficulty of predicting

outcome after severe brain damage have been dis-cussed previously in these columns. 1 When the

damage results from cardiorespiratory arrest, particu-larly if this has happened in the course of a hospitalprocedure, the problem may be complicated by theemotional involvement of the staff. If electro-

encephalography (E.E.G.) could provide a more reliableguide to outcome than does clinical assessment thiswould be a welcome development. Now a monographby PAMELA PRIOR 2 gives an invaluable review ofpublished reports on the E.E.G. in acute cerebralanoxia, as well as of data on 115 patients studied aftercardiorespiratory arrest in one hospital over 5 years.

’

Most were adults, so that the investigation comple-ments PAMPIGLIONE’S previous work in infants.3 The

- patients usually either made a complete recovery ordied a few days after their cardiorespiratory arrest;therefore the investigation was mainly concernedwith the role of the E.E.G. in predicting death orsurvival, rather than with forecasting quality ofsurvival. This sort of record demands good equip-ment and a skilled operator; and interpretation, evenof a satisfactory trace, presents considerable difficul-ties. PRIOR found that classification of the record

, into one of five grades by simple visual reading was

. of limited value: predictions were reliable only inmild or very severe cases, when the likely outcomewas usually clear on clinical grounds. In less obviouscircumstances greater accuracy was achieved by usingmore elaborate data-processing-either a linear dis-criminant score based on 36 variables, or an adaptivediscriminant score using 13 variables. Both thesemethods depend on visually interpreted aspects ofthe records, and, even with these refinements,predictions were apt to be overoptimistic. PRIORsees this as an error in the right direction, since itresulted only in renewed efforts on behalf of some

: patients who eventually died. This, however, is

precisely the dilemma for which a solution is urgentlyneeded, because uncertainty about the possibility of

,’, recovery still results too often in unreasonable..

prolongation of life-support. Not only is this need-lessly distressing to relatives and damaging to the

,

morale of nursing staff; it is also wasteful of resources.PRIOR recounts only 40 cases in detail, from which

it appears that an overoptimistic E.E.G. interpretation1. Lancet, 1973, i, 523.2. Prior, P. F. The EEG in Acute Cerebral Anoxia. Amsterdam, 1973.3. Pampiglione, G., Harden, A. Lancet, 1968, i, 1261.

was usually associated with clinical evidence ofirrecoverable brain damage, although that was nother conclusion for the series as a whole. In thosewhose outcome was accurately predicted (anddetails are published) the level of consciousnessand other clinical features likewise seem usually tohave indicated whether or not recovery was likely.How much can the E.E.G. add to what is clinicallyobvious soon after severe brain damage ? The

question is particularly important in the diagnosis ofbrain death-whether due to cardiorespiratory arrest,to head injury, or to other conditions-because thisproblem often arises in hospitals which do not haveE.E.G. departments. Certainly insistence that an

isoelectric E.E.G. is an essential criterion of brain deathis unfair to electroencephalographers, who may beasked to bear undue responsibility; it is unfair, also,to experienced clinicians, who may know from thewhole clinical picture that a patient will not recover,yet may be inhibited from acting accordingly. Thatis why the University of Minnesota report 4 waswelcome. It proposed criteria for brain death whichdepended on clinical observations only; it recom-mended that, when there was an irreparable lesion,E.E.G. was not only unnecessary but might lead to amistakenly optimistic prognosis. This paper repre-sented a logical development from the insistence on24 hours’ isoelectric E.E.G. in the 1968 Harvardcode which was later reduced to 12 hours in thesame hospital 6 and to 1 hour by the Cornell-affiliatedhospitals 7: already in 1969 BEECHER had suggestedthat an E.E.G. might not always be necessary. These

progressive codes have all come from groupsof doctors who were accustomed to workingtogether in a single institution. It is probablyunrealistic to expect committees with members from

widely different medico-social backgrounds to agreeon all aspects of such a delicate issue. Consequently,attempts to reach acceptable definitions of brain deathby national or international bodies seem likely eitherto fail altogether or to result in conclusions whichare so cautious as to be useless. Although no Britishinstitutional recommendation seems to have been

published, a code derived from the Minnesotaand Cornell criteria has been in use in one largecentre for the past 2 years. It applies only to

patients with structural brain damage in whomthere is no suspicion of depressant-drug intake orprimary hypothermia. On at least two occasions,12 hours apart, it must be shown that no spon-taneous respiration occurs during 3 minutes’ dis-connection from the ventilator, during which timeoxygenation is maintained by delivering 6 litres

oxygen per minute through a tracheal catheter.Provided arterial PC02 is normal (36-44 mm. Hg)4. Mohandas, A., Chon, S. N. J. Neurosurg. 1971, 35, 271.5. Harvard Medical School. J. Am. med. Ass. 1968, 205, 340.6. Beecher, H. K. New Engl. J. Med. 1969, 281, 1070.7. Plum, F., Posner, J. Diagnosis of Stupor and Coma. Philadelphia,

1972.