This paper presents a review of decision making in obstetrics. The first section deals with the obstetric problem-solving model, the second general decision theory and finally several aspects of decision making in the domain of obstetrics are examined.

Screening for genetic and congenital abnormalities during pregnancy has developed over the past 10–15 years from a concept to an established aspect of modern obstetric management in response to the declining importance of other causes of infant and perinatal mortality and morbidity. This review will consider why screening is indicated for this group of conditions, the criteria for any screening programme and the current impact of prenatal screening.

Human implantation has become the focus of considerable scientific interest with the advent of in vitro fertilization and other assisted conception techniques. The frustratingly high failure rates associated with these procedures has led to detailed appraisal of possible causes. Without basic knowledge of the physiological processes involved, little progress has been achieved. Investigations have broadly concentrated on the endometrium/uterine environment and blastocyst quality. However, information on both these aspects in the human being is limited. Even in animal reproduction, the mechanisms of implantation are far from understood, although some of the possible factors involved have been studied. By necessity, much of this review extrapolates from animal data.

The fetus has become increasingly accessible and visible as a patient over the last two decades. Two major technical advances have contributed to the obstetrician's ability to study and understand the fetus. The first advance was electronic fetal heart-rate monitoring, followed by the second major advancement in the form of real time ultrasound. Prior to the introduction of these technologies, biochemical determiniations of oestriol and human placental lactogen were used to monitor fetal wellbeing. Currently the emphasis, both in clinical practice and research protocols, is on biophysical surveillance of the fetus.

We cannot talk to the fetus. Our ability to place the fetal condition accurately in context, therefore, lies with indirect evidence and inferred responses. We can examine the fetus. We can do so in a most comprehensive fashion, addressing the salient points of fetal anatomy, both structural and functional. We are able to perform routine evaluation of every system and the methodology is available to assess critical areas such as the heart and the central nervous system in specific detail.

The puerperium is a time of substantial maternal risk. Within recent years, crude maternal mortality rates of 8.9/100,000 total births have been reported from England and Wales and of 6.6/100,000 from Sweden. Maternal deaths are considerably more likely to follow delivery than to precede it (Table 1). The risks associated with caesarean section are particularly noteworthy: death rates have been reported to be as much as 12 times greater than those following vaginal delivery.

There is an intimate relationship between the kidney and pregnancy. Renal plasma flow increases by 50–70% during a normal pregnancy and the glomerular filtration rate by about 50%.1 These changes commence in the first trimester and fall in the last trimester reaching normal levels within about four weeks postpartum. These physiological changes are accompanied by striking anatomical changes which consist of dilatation of the ureter, pelvis and calyces, together with an increase in renal parenchymal size. The dilatation i s more marked on the right and may appear in the first trimester. At term, 90% of pregnant women show this change.2

Sodium is the major cation in the extracellular fluid volume (ECFV) and as such, is the most important determinant of osmolality and of the volume of this fluid compartment. Hence any alteration in the control of body sodium will be reflected by changes in the ECFV, including the maternal plasma volume. There is no doubt that expansion of the plasma volume is a necessary and desirable event during pregnancy, influencing positively both maternal and fetal outcome. Therefore, studies of sodium balance in pregnancy provide important information relevant to both mother and fetus.

As a result of advances in perinatal and neonatal care, birth defects now represent an increasing component of the perinatal mortality rate. Approximately 5% of all pregnancies monitored result in an abnormal fetus and close to one-third of these are chromosomal abnormalities. The upsurge of interest in prenatal diagnosis has reflected this increased importance of genetic disorders in obstetrics.

During pregnancy and lactation there are many changes in maternal calcium physiology which maintain homeostasis in the face of greatly altered calcium balance. In the course of fetal growth and development, 30g of calcium is incorporated into the fetus by term, an amount derived wholly from the maternal system. Most of this accumulates in the latter half of pregnancy, representing a net transfer of 200mg calcium/day (5mmoles). The fact that this is not achieved at the expense of the maternal skeleton is testimony to the conservative and protective adjustments that are seen in calcium metabolism in pregnancy. Furthermore, the changes must persist both in the puerperium and later when lactation presents a source of continuing maternal calcium loss to the suckling infant. The calcium content of human breast milk s i 6–9mmols calcium/l, two to three times the maternal serum level. In the course of one week a normal breast-fed at term infant takes two to three litres of milk, containing 10–30mmols of calcium. The maternal daily calcium intake recommended by the World Health Organization s i 1.25g (30mmol) of which only 25% is absorbed. Thus calcium loss from mother to baby is significant and may not be replaced by diet in many parts of the world.

Pregnancy appears to increase the risk of thromboembolism sixfold, but there are wide variations in the reported incidence of thromboembolic complications during pregnancy. This is due to diagnostic difficulties and the inculsion of various clinical entities within the disease classification. However, there are reliable data on the incidence of fatal pulmonary thromboembolism (PTE) due to the detailed and thorough reports of the confidential enquiries into Maternal Deaths in England and Wales.

Efficient respiration in a baby relies on adequate anatomical development of the lungs, effective control of respiration, neuromuscular coordination and anatomical and biochemical maturation of the alveoli so that they are stable, thin-walled and in close proximity to the alveolar capillary bed. As well as maturation of lung function, there must also be a major change in pulmonary blood flow at birth to allow perfusion of the capillary bed with deoxygenated blood in order that effective gas exchange can take place. Successful initiation and sustaining of repiration after birth requires all of these things to occur simultaneously otherwise respiratory failure will ensue.

The established technique of investigating peripheral arteries by continuous-wave Doppler ultrasound was first applied to obstetrics by FitzGerald and Drumm in 1977. Since that time, each year has seen an increasing number of reports regarding obstetric Doppler, but assessment of its true value in pregnancy management, by large prospective randomized trials, is still awaited. Continuous-wave Doppler ultrasound is a relatively simple and inexpensive technique, ideal for screening the uterine and umbilical arterial circulations without the necessity for concurrent ultrasound imaging. However, difficulties in interpretation and analysis of the displayed Doppler-shifted frequencies, as well as the limited number of reports on the performance characteristics of Doppler ultrasound instruments, may well have contributed to the slowness with which the technique has been taken up for study. Continuous-wave Doppler systems designed for obstetric use have recently become available and the Vasoflo-3 (Oxford Sonicaid, Chichester, Sussex) has a microprocessor which makes the instrument very versatile. I shall review the technique of continuous-wave Doppler ultrasound measurement in obstetrics, with particular reference to measurement variation, in the light of our experience with the Vasoflo-3 (Figure 1). I shall then briefly summarize a few of the reports which have used continuous-wave Doppler ultrasound in obstetrics.