THE LANCET, MAY
22 1 1976
1137
aminoglycosides. The exclusion by repeated disc testing ,of organisms accords with Traub.' Certainly our in-vivo correlates of failure of gentamicin therapy confirm the in-vitro observations. 2 Inf(ctioll5 Disease Section J Research and Medical Services, Wadsworth Veterans Administration Hospital and School of Medicine, ' University of,California, , Los Angeles, Califoniia, U,S.A.
SIR,-Dr Willi~ms (May 8, p. 996) refen to data suaesting an association between alcohol consumption and an increased incidence of cancen of breast and thyroid and of malignant melanoma. It was suggested that alcohol might increase the occurrence of these malignancies by stimulating anterior-pituitary secretion of prolactin, thyrotropin, and melanocyt~timu lating hormone, and that this hypothesis should be tested. ALCOHOL CONSUMPTION AND DEATH-RATE FOR BREAST CANCER IN FOURTEEN COUNTRIES
Country
(liyr) France Italy Portugal' Austria ,",'est Gc'~~ny Australia Czechoslovakia Canada Iklgium United Kingdom Ireland Denmark Netherlands Finland
'2H 18·0 17·6 ' I4·S ,1306 10·7 10·3 9·0. 8·4 ' 7·7
Deathslyr/U)O 000 for malignant n
ofbmon .F
7·S 6·2 4·2
27·7 21·S 14-9 .. 28·9 28·9 23·9 ' 21·0 2S·1 3S·3 40·6' 28·6 36·2 34·8 18·0
Correiaiion coefficient{T)
= -0·32
7-6
N~tional Radiologic~1 ProteCiion IIoard, Harwell, Didrot, Oxfordshire OXH ORQ
M
0·6 0·3 (i·O 0·4 0·4 0·3 , 0·3 0·2, . 0·3
, . CJ.3
,
,0·1 0·1 0·7 0·1 +0·27,
The W.H.O. Expert Committee on Drug D~nden'ce cited estimates (collected by the Addiction Research Foundation, Toronto) of the'mean per-<:aput 'alcohol' c~nsuniption for fourteen countries.) The rates were expressed as litres o( a~ lute alcohol for perSons aged IS and older and ranged from 4·2 to 24· 7litres per year. '. ',' '. Mortality-rates fiom breast cancer for all fourteen countries , are given in the World Health Statistics Annual (1965 or 1968).4 The values are given in the table and show no significant correlation with estimated alcohol intake in theSe countries. Incidence-rates are cited in Cancer Incidence inFiW! Continents for. cancers of breast and thyroid in five' of these countries, and of malignant melanoma (of skin) in four of them.' These rather limited data again show no correlation significant at p=O·10. . Strictly these comparisons should presumably be made on the basis of an age-standardised population of over 1S' yean or from an age at which an alcohol intake might be supposed to have had an 'effect on cancer mortality or incidence. Such a comparison could be derived from the data but the lack of any 1. Traub, W. H. Appl. Mierobiol. 1970,10, 98. 2. Meyer, R. D., Lewi., R. P., Carmalt, E. D., Finegold, S. M. Ann. intern. Med.197S,81,484. 3. Tech. Rep. S". Wid Health OTK. 1974, no. SSI. 4. World Health Statistics Annuals 1965 and 1968. W.H.O., Gene"a, 1968 and , 1971. ' S: Doll, R., Muir, C., Waterhouse, J. (cditon). Cancer Incidence in Five Continents: vol. Il.lkrlin, 1970.
E. E. POCHIN
,A SCANDALOUS IMPROMPTU
RICHARD D. MEYER
ALCOHOL AND CANCER OF BREAST AND THYROID
Estimated alcohol, consumption per caput
significant correl~tion for any of th~ maliinanci~ in either sex is, however, discouraging (or perhaps encouraging).
'SlIt,:-you~ note on my piunphlet (May,1, p. 975) seems to have set out to destroy the credibility of a short but serious review by saying that'the argument is based on "questionable fact'l , "questionable assUmptions .... You use two exainplC$-'-as to facts; "that when ,the length of stay shortens the average' occupation is lowered", and as to assumptions that ,"single rooms in new hospitals presuppose 31% fewer~s". The fint is axiomatic in the Hospital Service and is demonstrated by the statistics, of all acute,h~pitals since 1948. It is not a questionable'fact, it is a proven and accepted Cact. The seCond is not an assumption but a d~uction from preceding statistics clearly demonstrated in the pamphlet, You then accuse.me of sleight of haild (i.e., dishonesty). That is both unfair and'misleading. , I could not and would not' object to opinions in a wholly adverse review but I hope I am entitled to object to 'a trivialising, by unsppported questioning of facts or figures, of the importance of propOsals for a significant advance in British hospital practice. Your note would lead your readers to suppose my proposals did not merit even casual thought.
or
RadeI,iKe Infirmary,
E. J. R. BURROUGH
Oxford OX2 6HE
RISING RATES OF CONGENITAL DISLOCATION OF '. "THEIDP?' ,
'SIR,-The hips of all infants born in this maternity hospital have examined clinically, twice in the fint week of life, in the past five yean 1971-75. The frequency of congenital dislocation of the hip (i.e., dislocated or dislocatable hips) has
been
been;
,
YeaT • 1971 1972 1973 " 1974 . 1975
,Neu>bor1l i,,/IUI/S 3143 ' 2S64 2360 2204 . 22S9
Inf4nlS ",irh CDR. (and Tate/1000) 18 (S·7) 14 (s.4) 21 (8·8) 17(7·7) , I S (6·6)
These findings de.> not accord with the rising rates of C.D.H. reported by Professor Klingberg and his colleagues (Feb. 7, " , p.298and Maich 6, p.547). , '.. , Maternity Hoopital, ~Ioniki,
Greece
ALEXANDER PEONIDE S
, ENDOMETRIAL CARCINOMA AND (ESTROGEN I THERAPY FOR BREAST,CANCER,
, Sut,-Dr Hoover and' his colleagues (April 24, p. 885), in re-evaluating the Possible association between endometrial carcinoma 'and ,
has