103
might attempt to rescue (i) by postulating, say, aI 30-year induction period in men and a zero inductionI We
period in women. This and similar alternatives can be: disproved by comparing the magnitudes of the increases inI SM and SF over the period 1920 to 1930 with sex-specific: cigarette consumption. 2,6 If we may assume that the sheerr scale of the increases in SM and SF render improbable any substantial contribution from
(iii),
we are
left with
(ii)
(iv) as possible explanations. We cannot readily dispose of (ii) except by showing that (iv) alone will suffice. In this connection, Rosenblatt et al.7 found that in a series of consecutive necropsies at the Doctor’s Hospital, New York, 1960-1969, lung cancer constituted 5-5% of all malignancies. This proportion was markedly lower than the corresponding values found in late 19th-century and early 20th-century series from Austria, Germany, and the U.S.A., in which necropsy was the chief method of diagnosis. In these early necropsy series (five in all) lung-cancer diagnoses ranged from 8-3% to 11-5% of all cancers.’ A strict comparison between these early and the recentstudies would call for the control of histological It would, therefore, be as criteria, country, age, sex, &c.
been followed-up over a sufficiently long time for Dr Maddison to be sure that the incidence of subsequent malignancy in these patients is not significantly increased ? The inclusion of a progestogen in the cycle does not necessarily reduce the dangers of prolonged hyperarstrinism in the older age-groups. 76 Hill Top Avenue, Cheadle Hulme,
Cheadle,
and
unwise
conclude from the direct necropsy evidence that rates have genuinely decreased during this century as it would be to conclude from national mortality statistics that they have genuinely increased. Mr Cole (July 7, p. 49) misses a key point. Although the shapes of the age patterns3 recorded for lung cancer in men have remained surprisingly similar from period to period and from country to country, it will be seen from fig. 1 of reference 3 that the modal age in the England and Wales statistics climbed steadily from t= 65 yr. in 1951-55 to t= 73 yr. in 1966-70. This is a consequence of the exceptionally large secular increases in recorded death-rates in the highest age-groups.41 suggested3that this divergence of the statistics from an equation with constant k could be accounted for " in terms of an artefact produced by falsepositive diagnoses ". Mr Cole does not indicate whether he is prepared to attribute the factor of 58 increase in recorded death-rates at 80-84 years of age, over the period to
lung-cancer
1921-25 to 1960-61, to increased cigarette consumption. As we have seen, the reality of this and other rises in deathrates is not supported by necropsy evidence.’ Despite problems of intercomparison the necropsy evidence for secular trends is the most reliable we have. In conjunction with figs. 1 and 2, it shows that the vast increases recorded in death-rates from lung cancer during this century are largely or wholly the result, not of Professor Smith’s anonymous agent ", but of diagnostic artefacts. With little or no genuine increase to explain, we cannot seriously pretend " that cigarette smoking causes nearly all cases of lung cancer ". But entrenched myths die hard and mind-changing can be painful. "
General Infirmary, Leeds LS1 3EX.
P. R. J. BURCH.
HORMONAL REPLACEMENT THERAPY FOR MENOPAUSAL SYMPTOMS
SiR,—Iread with interest the letter from Dr Maddison (June 30, p. 1507) about hormone replacement therapy for menopausal symptoms. I am sure that many menopausal women may obtain subjective benefit from such medication, but Dr Maddison omits to say how long the treatment continues.
I
was a little surprised that " in many cases been continuous over 10 years ", and I am rather concerned at the possible effects on the breast and endometrial tissue in such patients. Have enough patients
therapy has
7.
Rosenblatt, M. B., Teng, P., Kerpe, S., Beck, I. Medical Counterpoint, 1971, 3, 53.
Cheshire.
IAN J. KERBY.
SAFE RESUSCITATION AT BIRTH
SIR,-Fears have often been expressed that resuscitation for intrapartum asphyxia by endotracheal intubation and intermittent positive-pressure ventilation at birth must increase the risk of pneumothorax and pneumomediastinum, 1,2 but there have been very few reports of trouble. The discovery of a large pneumothorax at post mortem in two severely asphyxiated babies of 41 weeks’ gestation who died during attempts at resuscitation immediately after birth therefore prompted us to examine the inflation pressure generated by the equipment that was in use. Neither of the babies breathed spontaneously after birth. The babies were not malformed and their lungs appeared macroscopically normal, but there was amniotic debris in the terminal airways, and this was particularly striking in the child who died about 10 minutes after birth with a massive right-sided tension pneumothorax. A spring-loaded safety valve is employed to limit inflation pressure in some resuscitation equipment,3 and an adjustable autoclavable spring-valve4 is now on commercial sale in this country, but a simple water manometer of the type popularised by Hamer Hodges5 and Barries is used to monitor and limit pressure in the Vickers Medical resuscitation trolley that has been on the market for over 10 years and is now in general use in this country. Examination of the trolley that was used to resuscitate the two babies who died showed that gas was being delivered at a rate of 10 litres per min. and a static pressure of 100 cm. H2O when the flow control was fully open, and this led us to examine a further 10 trolleys in local use. Maximum flow varied between 10 and 12 litres per min. and maximum pressure varied between 43 and 115 cm. H2O; 7 of the 11 trolleys generated a maximum pressure of more than 50 cm.
H2O.
Mathiasdrew attention to the limitations inherent in the use of a water manometer to control inflation pressure in 1966, but the variable performance of commercial equipment has not so far been the subject of comment. Our own findings are summarised in the accompanying figure; there is a quite unexpected variation in the performance of the 11 trolleys we tested that we are at a loss to explain. We are reminded of the sudden discovery by Hustead and Averythat the pressure generated by the Kreiselman9 resuscitation apparatus used so widely for nearly thirty years in America was, in fact, extremely variable and often greatly in excess of the pressure indicated by the water manometer. The price of safety is eternal
vigilance. Cook, C. D., Lucy, J. F., Drorbaugh, J. E., Segal, S., Sutherland, J. M., Smith, C. A. New Engl. J. Med. 1956, 254, 604, 651. 2. Hutchison, J. H., Kerr, M. M., Williams, K. G., Hopkinson, W. I. Lancet, 1963, ii, 1019. 3. Saling, E. Geburtsh. Frauenheilk. 1958, 18, 128. 4. Goddard, P., Becket, A. J. Lancet, 1971, ii, 584. 5. Hamer Hodges, R. J., Tunstall, M. E., Knight, R. F., Wilson, E. J. Br. J. Anœsth. 1960, 32, 9. 6. Barrie, H. Lancet, 1963, i, 650. 7. Mathias, J. ibid. 1966, i, 262. 8. Hustead, R. F., Avery, M. E. New Engl. J. Med. 1961, 265, 939. 9. Kreiselman, J., Kane, H. F., Swope, R. B. Am. J. Obstet. Gynec. 1928, 15, 552. 1.
104 The manufacturers intend to check the safety of all the that they service and maintain on contract to ensure that no trolley is capable of generating a pressure of more than 50 cm. H2O; they are also introducing a suitable warning label which will be affixed next to the flow-meter, An independent dead-weight relief valve (at 30 or, optionally, 40 cm. HO) is being incorporated into the manometer circuit of all new equipment to eliminate any residual hazard, and this further modification will be added to equipment already in service upon request.
equipment
Princess Mary Maternity Hospital, Great North Road, Newcastle upon Tyne NE2 3BD.
Relation between gas-flow and static inflation pressure in 11 commercial resuscitation trolleys provided with a pressurelimiting water manometer 30 cm. high when flow regulator was set at 1, 2, and 3.
Estimates of maximum safe inflation pressure vary
greatly. Figures of as little as 25 cm. H20 are still being quoted,lo but pressures of up to 40 cm. H20 are now usually considered acceptable, and post-mortem studies 11 seem to indicate that normal lungs are unlikely to rupture until pressure exceeds 50 cm. H2O. A static pressure of 30 cm. H2O is enough to expand the lungs of most healthy babies at birth if maintained for about a second, even when inflation fails to provoke a gasp,l2 and pressures of 20 cm. H20 are usually enough to maintain adequate ventilation once aeration is achieved. However, the lungs of some small stillbirths have an abnormally high opening pressure,13 while the lungs of many of the babies who subsequently develop respiratory distress have an abnormally low compliance at birth 14; these sometimes require a higher inflation pressure during resuscitation. A gas flow of 2 litres per min. is generally adequate when resuscitating a baby who has been intubated, though inspiratory flow may momentarily exceed 10 litres per min. when a baby makes a sudden inspiratory gasp,15 and precautions are therefore necessary to ensure that water can never be drawn into the lungs; flows in excess of 2 litres per min. may, however, be necessary to combat leaks when an anaathetic gas bag and a face-mask are used to initiate resuscitation, and Vickers Medical therefore make provision for this. Unfortunately, it is not generally appreciated that the flow-meter provided with this equipment is calibrated in arbitrary units and that flows in excess of 5 litres per min. will be obtained when the flow-meter is set at two. We feel that all who use this efficient and convenient resuscitation trolley should be made aware of the very variable and sometimes dangerously high pressure that can be generated by high gas flow. It should be stressed, however, that there is no hazard as long as gas flow is kept below 4 litres per min., a figure that is more than adequate for a baby who has been intubated at birth. 10. 11. 12. 13. 14. 15.
Brown, R. J. K., Valman, H. B. Practical Neonatal Paediatrics; p. 9. Oxford, 1971. Rosen, M., Laurence, K. M. Lancet, 1965, ii, 721. Hull, D. J. Pediat. 1969, 75, 47. Gruenwald, P. Lab. Invest. 1963, 12, 563. Hey, E., Hull, D. J. Obstet. Gynœc. Br. Commonw. 1971, 78, 1137. Long, E. C., Hull, W. E. Pediatrics, 1961, 27, 373.
EDMUND HEY WARREN LENNEY.
HOSPITAL ADMISSION AFTER HEAD INJURY SIR,-Mr Potter (June 16, p. 1381) advances several reasons why patients without fracture should be admitted after head injury involving brief amnesia. It is the possibility of the development of extradural haematoma that is uppermost in the minds of most general and orthopxdic surgeons, to whose wards these patients are at present admitted in such large numbers, and my paper was directed towards that complication only. I was careful to stress that adequate X-ray facilities were required; with improved accident services being developed in this country these are more often available than they used to be, and it seemed worth emphasising one benefit that could arise from such facilities. In any event, there are many cases in which fractures can be reasonably excluded. Medicine being an inexact science based on reasoned judgments, it is impossible to legislate for everything; surely clinical judgment should continue to be made on medical rather than on medicolegal grounds. The easy decision is no decision-that is to admit all cases. There is reason to doubt whether the present policy is the most effective, because the admission of so many mild cases which do not suffer complications tends to blunt people’s awareness of the likelihood of such events. If fewer cases were admitted on a more selective basis they would almost certainly be better managed, and there might even be time for the careful reassurance and assessment which Mr Potter calls for. Institute of Neurological
Sciences, Southern General Hospital, Glasgow G51 4TF.
SAMUEL L. GALBRAITH.
HYPERGLOBULINÆMIA IN LIVER DISEASE SIR,-We should like to comment on some of the points made by Dr Triger and Professor Wright (June 30, p. 1494). Their speculation that defective sequestration in Kupffer cells of a wide range of antigens occurs in liver disease has found experimental support in our findings1 that cirrhotic rats clear particulate antigens and immune complexes from the portal blood less efficiently than normal rats. Moreover, these observations are consistent with Triger and Wright’s finding that liver damage annuls the difference in antibody responses to sheep erythrocytes between portal venous and inferior caval routes of injection. Our studies with induced hepatic necrosis showed that
these rats developed hyperglobulinsemia of equal magnitude to that found in cirrhotic rats. Thus, although portocaval shunting may contribute as a factor in the failure of the diseased liver to sequester antigens absorbed from the gut, damage of Kupffer cells or their blockade by products of cell damage can produce the same result.2 Thomas, H. C., MacSween, R. N. M., White, R. G. Lancet, 1973, i, 1288. 2. Horne, C. H. W., Thompson, W. D., Busuttil, A., MacSween, R. N. M. Br. J. exp. Path. 1973, 54, 222. 1.