Down's syndrome (DS) is the commonest cause of severe mental retardation in children. It is the result of trisomy of chromosome 21 which is usually a random event though it is commoner in older mothers. DS can be diagnosed by chorionic villus sampling (CVS) and amniocentesis followed by karyotyping. Because of the risks associated with these invasive procedures, they can only be offered to a high-risk group. At one time the sole basis for identifying this increased risk was maternal age, but within the past ten years a series of biochemical and ultrasound abnormalities have been shown in DS pregnancies. The biochemical abnormalities include changes in the levels of most fetal and placental products in the maternal circulation. The best-known of these changes are the reduced levels of alphafetoprotein (AFP) and oestriol (E3) and increased levels of human chorionic gonadotrophin (hCG). The mechanism underlying these biochemical phenomena is unknown. Screening programmes involving the measurement of hCG and AFP, with or without additional parameters such as E3, at 15-18 weeks of pregnancy can typically identify 60% or more of cases of DS with a screen-positive rate of 5%. The combined risk derived from the various biochemical parameters, together with maternal age, is calculated by one of a number of computer programmes which have been developed for this purpose. There has been considerable discussion as to the exact biochemical tests which should be used for DS screening. This had led to controversy as to whether measurement of E3 has a place, and whether or not measurement of the free beta-subunit of hCG should replace measurement of the intact molecule. A notable recent development is the suggestion that measurement of the urinary beta-core of the hCG could be a highly discriminatory marker. A number of factors can affect the results of biochemical screening for DS. These include maternal weight, gestational age, ethnic origin, smoking, and diabetes. In addition, abnormal levels of the biochemical products may be found in other chromosome abnormalities.

facilities serving the clinics are limited to Northwick Park Hospital and to one other site, which provides services to the patients of general practitioners but is staffed and organised by the Northwick Park radiology department. Five patients out of 491 referred for radiography had a repeat x ray examination because of missing films: a repeat radiography rate of 1%. Furthermore, the Scottish authors expressed their 33% of repeat examinations as a percentage of all patients attending the orthopaedic clinic (whether sent for radiography or not)4; our equivalent figure for repeat radiography for all patients attending our orthopaedic clinics was substantially less than 1%. The important message from the Scottish survey would seem to be that health centres that have x ray facilities may be associated with a high level of repeat x ray examinations in patients subsequently referred to hospital for an orthopaedic opinion. Clearly, that particular problem needs to be dealt with. For understandable reasons of brevity the BMJ editorial failed to convey the particular background to the Scottish survey.4 When BMJ7 editorials, commentaries in the lay press,' or indeed review articles such as Fiona Godlee's refer to unnecessary x irradiation in this particular context the authors might consider that it could be reassuring to their readers, as well as to patients, to note that a repeat rate of less than 1% has been shown to occur in the setting of a district hospital where local health centres do not have x ray facilities. At present this represents a more typical setting for the United Kingdom in general than do the arrangements existing in Scotland.