J. R. Gibson

Observations on All Births (23,970) in Birmingham, 1947

The short clinical histories and the diagrams show that a distinct remission occurred in at least five out of six polycythaemic patients in response to the oral aiministration of /3-naphthyl-di-2-chloroethylamine for periods varying from a few weeks up to a few months. At first we were cautious with the dosage; 500-600 mg. might be given daily to start with for a few months, after which the dose should be reduced. An attempt was made to keep two patients stable on a small continuous dose (200 mg. daily), so far with good result. We have also seen a fall in the leucocyte values during the treatment; in one case a fall to as low as 1,200 leucocytes pzr c.mm. occurred, but a spontaneous and rapid rise set in 12-14 days after the drug was discontinued. It appears from this and what was earlier stated that the substance is a potential bone-marrow poison inherent in which is the risk of damage to the bone marrow, with aplastic anaemia or perhaps rather granulocytosis as a result. It is therefore necessary that these patients should be carefully and constantly followed with blood counts during treatment. As is known, polycythaemia is in itself a relatively benign disease and usually the patients have a fairly long expectation of life (cf. Dameshek (l950) and VIdebaek (1950), amongst whose groups of patients there were several with a history of over 20 years), and therefore emphasis must be placed on caution during the treatment.

FOETAL AND INFANT MORTALITY IN MULTIPLE PREGNANCY

INTRODUCTION IT is well established that risk of stillbirth is greater for twins than for single births, and greater for triplets than for twins. Using national statistics, Yerushalmy and Sheerar (1940), Strandskov and Ondina (1947), and Lowe and Record (1951) showed that the stillbirth rate is inversely related to the number of foetuses in the uterus, and many workers have recorded high stillbirth rates in twins born in hospital (Neuhauser, 1914; Marinoff, 1926; Takahashi, 1934; McClure, 1937; Gernez and Omez, 1938; Hirst, 1939; Munnell and Taylor, 1946; Vermelin and Ribon, 1948). Mortality rates of multiple live births are less reliable. National statistics are not published in a form suitable for calculation of neonatal or infant mortality rates for multiple births, and available data are derived mainly from hospital experience. Few direct comparisons have been made between twin and single births; in general, recorded twin neonatal mortality rates appear to be high (Neuhauser, 1914; Marinoff, 1926; Takahashi, 1934; Hirst, 1939; Potter and Crunden, 1941; Potter and Fuller, 1949). There are no satisfactory records of subsequent mortaIity in twin and single births which survive until the end of the first month. So far as we are aware, the only information about mortality in liveborn triplets was provided by Fisher (1928), who reported that after 146 triplet deliveries, 39 per cent of male and 46 per cent of female infants were alive at 64 years. We know of no data on liveborn quadruplets. The higher mortality of twins than of single births has been variously ascribed to toxaemia, 47 1 crowding in the uterus, complications of delivery, and prematurity. (1) Toxaemia is a more frequent complication of twin pregnancies than of single pregnancies (Guttmacher, 1939; Potter and Crunden, 1941; Munnell and Taylor, 1946; Vermelin and Ribon, 1948), but there is little direct evidence that it contributes to the higher mortality of twins. Indeed figures quoted by Strandskov and Ondina (1947) suggest that toxaemia is not responsible for the discrepancy between mortality rates of twin and single births. (2) Crowding in the uterus is regarded by Newman (1940) as the main cause of high prenatal mortality in twins. The deaths of some monozygous twins are undoubtedly due to circulatory imbalance resulting from the placental anastomosis (Price, 1950); foetal mortality is higher for twins of like sex than for twins of unlike sex (Yerushalmy and Sheerar, 1940; Lowe and Record, 1951), and for monochorial twins than for bichorial twins (Vermelin and Ribon, 1948). The observation that monochorial twin pregnancies terminate at an earlier stage than bichorial pregnancies also has a bearing on this problem (Colloridi, 1935; Mazzullo, 1948). (3) Complications of delivery. Munnell and Taylor (1946) reported an increased incidence in multiple births of uterine inertia, hydramnios, prolapse of the cord, and placenta praevia. It is commonly believed that these and other complications (premature separation of a normally situated placenta, malpresentation) are more frequent in twin deliveries, but it is not certain to what extent they are responsible for the in-

Observations on all Births (23,970) in Birmingham, 1947: II. Birth Weight

Observations previously published on the duration of gestation (Gibson and McKeown, 1950) and birth weight (McKeown and Gibson, 1951) lend themselves to an examination of the association between these variables and foetal mortality. Data available for single births in Birmingham during 1947 are exhibited in Table I, separately for stillbirths, first-year deaths, and births which survived to the end of the first year of life. We have previously discussed the representativeness of these records, and need here note only that there are no conspicuous differences between the proportions complete in each of the three classes.

Observations on all Births (23,970) in Birmingham, 1947. VI. Birth Weight, Duration of Gestation, and Survival related to Sex.

BIRTH WEIGHT AND DURATION OF GESTATION That mean birth weight is higher for males than for females has been consistently recorded (for example by Pearson, 1900; Murray, 1924; Martin, 1931; Bakwin and Bakwin, 1934; Anderson, Brown, and Lyon, 1943), and more recently by Karn and Penrose (1951), Norval, Kennedy, and Berkson (1951), and Salber and Bradshaw (1951). There have been fewer investigations of duration of gestation; but in general reported differences between the two sexes have been trivial (Schlichting, 1880; Siegel, 1921; Anderson, Brown and Lyon, 1943; Karn, 1947; Karn and Penrose, 1951). Table I gives mean birth weights of the Birmingham births as 7 57 and 7 31 lb.

Observations on all births (23, 970) in Birmingham, 1947. VII. Effect of changing family size on infant mortality.

The object of this communication is to enquire to what extent the reduction in infant mortality during the 20th century may be attributed to changes in family size. This important question was referred to in Volume XIII (Pt II) of the 1911 Census of England and Wales (p. xlix). The census recorded for each married woman the total number of children born alive, and the number living and dead at the time of the Census. This made it possible to relate mortality to family size (but not to order of birth), and it was noted that: Whatever the ages of the parents and the duration of marriage the increase of mortality with number of births is very great, the rates being generally three to five times as high for the largest families as for the smallest.