695 F. Forman for his continued interest and encouragement. The Naffziger operations were performed by Mr. de Villiers Hamman and Mr. A. Gonsky. The expenses incurred in connection with this report are defrayed by a grant from the Council for Scientific and Industrial Research. Endocrine Clinic, R. HOFFENBERG Department of Medicine, Groote Schuur Hospital and P. U. JACKSON. W. University of Cape Town.
fluid was infused. No ill effects were seen; the blood-pressure fell in four cases, but not dangerously. Four patients recovered and were able to eat. Two later died, one (who had had a hasmatemesis before treatment) of bronchopneumonia, and the other with a recurrence of status epilepticus.
All these patients were seriously ill, dehydrated, and difficult to nurse. During sedation with hydroxydione for varying periods, very necessary fluid was given with
ORDER AND DISORDER IN THE CEREBRAL CIRCULATION
comparative
SIR,-Sir Russell Brain1 quotes the investigations of Lawrence and Rewell2 into the cerebral circulation of the giraffe. These workers suggested that in the giraffe blood might be shunted from the carotid artery into the vertebral artery, with which the carotid communicates freely, and that this shunt would protect the brain of the giraffe from excessive hydrostatic blood-pressures. We have recently published our findings 3 in this species, and have shown that the rete mirabile is supplied by the vertebral artery. The connection was demonstrated by dissection and by angiography. If these findings are accepted, then Lawrence and Rewell’s hypothesis falls away. It was in any case difficult to understand why blood should shunt from one large artery into another, in which the blood-pressure would presumably be similar. There is no proof of any shunt mechanism in this animal. To make their suggestion’more plausible, Lawrence and Rewell had postulated that the giraffe would be found to have a high blood-viscosity. This has also not been verified in previous observations from this laboratory.44 We have suggested3 an alternative means by which the cerebral circulation of the giraffe may be adapted to the gravitational effects caused by changes of posture in this animal. R. H. GOETZ E. N. KEEN. University of Cape Town. HYDROXYDIONE IN DEHYDRATED PSYCHOTIC PATIENTS
SiR,—Since hydroxydione (’ Viadril ’) seemed a very drug (its L.D’50 is three times that of thiopentone), one of us (D. E. M.) suggested its use as a general sedative in disturbed patients; since it must be given intravenously in dilute solution, disturbed dehydrated psychotic patients seemed particularly suitable. safe
2-5 g. of hydroxydione was dissolved in 500 ml. of sterile 5% glucose saline solution. Two drip sets were used. A glucose saline drip was started, and the needle of a hydroxydione
drip was inserted into
the lumen of the lower part of the rubber
tubing of the first drip and secured by adhesive. Each drip was controlled independently by a screw clip. This ensured that (1) the transfusion channel was functioning properly before the hydroxydione was used, thus diminishing the risk of tissue necrosis; (2) the solutions could be used independently or together; and (3) the transfusion channel would not become blocked on change-over from one solution to the other. The hydroxydione solution was given at 150 drops a minute until the patient fell asleep; it was then turned off and the glucose saline drip turned on. The hydroxydione was restarted at the same rate when the patient began to show signs of restlessness, until he again fell asleep. Six patients were treated-three with senile psychoses, one with paraphrenia, one with status epilepticus, and one with delirium tremens. All were successfully kept asleep for periods ranging from 2 hours to 3 days, during which 1-7 litres of 1. 2. 3. 4.
Brain, R. Lancet, 1957, ii, 859. Lawrence, W. E., Rewell, R. E. Proc. Zool. Soc., Lond. 1948, 118, 202. Goetz, R. H., Keen, E. N. Angiology, 1957, 8, 542. Goetz, R. H., Budtz-Olsen, O. S. Afr. med. J. 1955, 29, 773.
ease.
r
,-.
"
___
D. E. MUNRO D. A. KNOX.
Stanley Royd Hospital, Wakefield.
__
_
AND MALIGNANT DISEASE suggestion of Dr. Alice Stewart and’her
RADIATION, LEUKÆMIA,
SIR, The colleagues1 that antenatal X-ray examinations may cause malignant disease and leukxmia in the child has been ably refuted by your correspondents,2-s but the growing publicity that this hypothesis has obtained and the widely expressed fear of diagnostic irradiation that it has encouraged demand further action. This should be your duty because of the uncritical enthusiasm with which you endorsed the preliminary communication in a leading editorial.11 Your editorials are so influential that no in your correspondence columns can outargument them with most readers, and although you are not weigh to act as a referee in all such arguments, in this expected of because case, your original attitude, you are morally bound to express a further opinion. When you do so may I suggest that you emphasise the following points ? Stewart et al. have agreed that " Dr. Reid is justified in stressing that, on the evidence of our figures, over 90% of cases of malignant disease in children under 10 years of age (which are in any case rare) must be due to causes other than diagnostic irradiation ". In the less than 10% of cases in which the children were irradiated antenatally these other causes could not be excluded, and the probability is that the same incidence of " other causes " would have been operative in this group. The absurdity of postulating diagnostic irradiation as a possible cause becomes manifest when this argument is applied to the
remaining less-than-1 %. The controls used in this work were inadequate.2-5 8 Stewart et al. did not deny their inadequacy; instead they stated 7"as previously reported, a higher incidence of antenatal irradiation is still found among our cases when comparison is restricted to controls of corresponding parity ". Unfortunately, reference to the letter they 9 does not appear to substantiate this statement, quote for therein appears: " they were deliberately not matched for birth rank ", and " I am not at the moment in a position to quote parity figures for the mothers included in the preliminary report"; and the controls remain nebulous and inadequate. The figures given by Stewart et al. do not show, as they claim, " that the child population dying of leukxmia and cancer
nowadays includes
a
disproportionately large
number of first-born and twins " ; nor is there evidence that matching for birth rank would have obscured this information if it were present. Indeed, such matching would have provided a sounder basis for attempting to 1. 2. 3. 4. 5. 6. 7. 8. 9.
Stewart, A., Webb, J., Giles, D., Hewitt, D. Lancet, 1956, ii, 447. Rabinowitch, J. ibid. p. 1261. Rabinowitch, J. ibid. 1957, i, 219. Reid, F. ibid. p. 428. Sonnenblick, B. P. ibid. p. 1197. ibid. 1956, ii, 449. Stewart, A., Webb, J., Giles, D., Hewitt, D. ibid. 1957, i, 528. Ellis, F., Lewis, C. L. ibid. 1956, ii, 573. Stewart, A. ibid. p. 573.
696
evaluate the part played by diagnostic antenatal irradiation than Stewart’s inability to think " why, when first conceived, first-born children should be more likely to develop leukaemia or cancer than other children" .9 There was a time when people could think of no other reason than the evil eye. The suggestion that individual twins received significantly different doses of antenatal irradiation is baseless speculation. In the 10 years under review an antenatal X-ray examination often included several views-e.g., anteroposterior, posteroanterior, lateral, and both A-P and P-A half-axial (Thom’s and Chassard-Lapine’s) views-and my guess (as good as Stewart et al.’s) is that the individual twins would not have received greatly different doses. What do Stewart et al. mean by " a disproportionately large number of first-born and twins " ? In proportion to what ? Did they consider that the risk of malignancy per pregnancy must be at least twice as high for twins as for singletons because there are twice as many children per pregnancy ? In any case, even if we accept that these diseases and antenatal X-ray examinations are commoner in twins and in the first-born’it does not follow that there is a causal relationship. The post-hoc variety of pretentious nonsense Huff 10 warns: is another way of changing the subject without seeming to. The change of something with something else is presented as because of." This post-hoc-propter-hoc argument is particularly inapplicable in a case such as this, where diagnostic irradiation is extremely common and the alleged sequel extremely rare; and furthermore, where the same disease is known to occur as the result of other causes in at least 90% of cases. "
Apart from the unsatisfactory method of the selection of the controls, the method of the collection of the information is open to question, and there is no evidence that the object of the interviews was hidden from the interviewing doctor or from the subject interviewed. Moreover, a slight bias could be introduced by the fact that a mother of a child dead of malignancy would have recalled the incidents of that particular pregnancy many times, and would only too readily catch the idea that an X-ray examination may have been responsible. The incidents of a first pregnancy would also be more readily recalled. But the incidents of later pregnancies with normal children would not be as likely to stimulate the memory. This bias, however, may not have affected the results materially, and in view of the fundamental fact of the inadequacy of the control group, it may not even be important at all. Perhaps more important is the possibility of unconscious bias in the authors. This one cannot fail to suspect. Mr. Hewitt has already contributed to the study of leukxmia in relation to ankylosing spondylitis and X-ray therapy. Without minimising this achievement, one cannot but wonder whether in the work being discussed here he merely found what he sought. The ease with which the authors attribute leukaemia and malignant disease to X-ray examinations because they can think of no other reason, their unquestioning acceptance of the significance of the word " routine " on an X-ray request form, and the important apparent contradiction (noted above) in their evidence also seem to point to bias. The work of Stewart et al. has assumed the importance that you attributed to it as " the first published epidemological evidence of the hazards of diagnostic radiology to the patient ".6 As such it deserves re-valuation by to other speculative extrapolations without reference you from massive doses of irradiation in man and in experimental animals, for there is no other relevant suggestion 10.
Huff, D. How
to
Lie with
Statistics; p. 135. London,
1954.
compared with the doses obtaining in diagnostic procedures. You would be rendering a service to the profession and to the public if you would persuade Dr. Stewart and her colleagues to publish the full results of their survey, of which we have had, after 18 months and in spite of its alleged importance and wide publicity, only the preliminary report. HARRIS JACKSON. Johannesburg.
that these
can
be
regrets about publishing the preliminary communication by Dr. Stewart and her colleagues: we see it as one of our functions to make known as soon as possible those findings which, we judge, may turn out to be of crucial importance. By our remarks at the time we sought to draw attention to the possible significance of these preliminary results: we did not endorse any particular conclusion that might be drawn from them. We have shown Dr. Jackson’s letter to Dr. Stewart and we are glad to learn from her reply, which follows, that the final report on the survey is nearly complete. We await with interest an opportunity to reassess the situation.—ED. L.
’*’ ’*’ ’*’ We have
no
SIR,-Like Dr. Jackson, we regret the length of time which has elapsed between our preliminary communication and the appearance of a full report. This delay has resulted from the unusually large scale of the survey and the fact that we have studied a number of other topics besides fcetal irradiation. In the meantime we have always given the fullest available information to anyone who was sufficiently interested to write to us direct. Dr. Jackson’s criticisms are of the most general kind and cannot be answered adequately in a letter. In our final report we shall deal with the possibilities of error arising from imperfections in the selection of controls and from bias in the mothers’ reports. As this report is now nearly complete and should be in print shortly, we would ask readers who share Dr. Jackson’s scepticism to suspend judgment until they have seen the evidence in full. ALICE STEWART. D. HEWITT. J. W. WEBB.
Department of Social Medicine, Oxford
University.
THE DOCTOR’S ATTITUDE TO HIS PATIENT
SiR,-Had Dr. Ryle, in his letter last week, been able to
personality (it would have been too soon to have expected a change) he would not have revealed himself as such an unscientific observer by attributing statements to me which I never made. The point at issue is this. In the past doctors have not thought it worth while to investigate functional disorders, partly because the scientific instruments in use always gave negative results. Today the position has changed; these disorders are being subjected to scientific inquiry and the main instrument used is the psychiatrist’s own mind. Just as the scientist recognises and allows for the errors of the scientific instruments he has fashioned, so the psychiatrist must recognise and allow for the errors of his own mind
suppress part of his
if he attempts
to use
it
as
a
scientific instrument. In
recognise the errors of his mind, the psychiatrist undergoes some personality change; for, in seeing himself at times as others see him, he becomes less bound by his previous pattern of behaviour and more receptive to the behaviour problems of his patients. Scientific inquiry has shown that the psychoanalyst cannot practise his specialty without undergoing an learning
to