British Journal of Obstetrics and Gynaecology June 1992, Vol. 99, pp. 527-532 CORRESPONDENCE zyxwvu Doppler investigation of uteroplacental blood flow resistance in the second trimester: a screening study for pre-eclampsia and intrauterine growth retardation Dear Sir, Bewley zyxwvutsrqpon ef al. (1991) underestimate the potential value of uterine artery Doppler studies as a screening test for abnormal pregnancy out- come. Their definition of an abnormal waveform is too rigid and the endpoints of adverse pregnancy outcome are incomplete. A review of the published literature on the outcome of pregnancies complicated by reduced uteroplacental perfusion reveals the following four abnormal outcomes: Pre-eclampsia. Robertson et a/. zyxwvutsrqp (1 967) first described the pathological lesions of the spiral arteries in pre-eclampsia. Steel et a/. (1990) found that those patients with a single abnormal uterine artery waveform, showing a high resistance pattern at 24 weeks gestation had a signifi- cantly higher risk of hypertension in later pregnancy than those with normal waveforms. Placental abruption. This is more likely to complicate a pregnancy with pre-eclampsia (Abdella etal. 1984). Uzan etal. (1 989) found that abnormal uterine artery Doppler flow waveforms preceded placental abruption by 34 weeks in the majority of their patients. Intrauterine growth retardation (IUGR). The same vascular disease of the placental bed that occurs in pre-eclampsia is seen in pregnancies with IUGR without pre-eclampsia (DeWolf et al. 1980). Steel et a/. (1990) found that patients with an abnormal uterine artery waveform at 24 weeks gestation had increased risk of an infant with asym- metrical IUGR. Prrterm delivery. Naeye (1989), in a study of placentas from 5414 pre- term deliveries, found that 26% showed evidence of compromise of uteroplacental blood supply. Although most came from pregnancies complicated by pre-eclampsia, 38% were not, suggesting that preterm delivery may be the zyxwvutsrqp only abnormal outcome of a pregnancy com- plicated by reduced uteroplacental blood flow. Steel et al. (1990) found that those patients with a single abnormal uterine artery wave- form at 24 weeks gestation had an increased risk of preterm delivery. Many studies still use birthweight for gestational age below the tenth or third centile as their definition of IUGR. Altman & Hytten (1989) expressed disquiet about this definition. Miller & Hassanein (1971) used Rohrer’s ponderal index (PI) [weight (g) zyxwvuts x 100/length (cm)’] to define asymmetrial IUGR, and found that birthweight by itself is not a valid measure of fetal growth impairment. In our study of 2508 consecutive births, we found that infants with a low ponderal index had higher rates of fetal distress and caesarean section delivery than infants with low birthweight (Fay et a/. 1991). As previously reported, many of the infants with a low ponderal index and a poor obstetric outcome actually had a birthweight in the normal range (Villa et al. 1990). Methods of early detection and management of disease can only be assessed properly if accurate endpoints are used. Bewley zyxwvutsr et a/. (1991) define outcome as: birthweight < 10th centile, or PIH, or APH of any cause, or intrauterine death of any cause or an operative delivery for fetal distress. Severe outcome is defined as: birthweight <3rd centile, or severe proteinuric hypertension, or placental abruption, or intra- uterine death of any cause. The former definition is inadequate and the latter incomplete. Without taking into consideration all the possible outcomes, there will be false positive tests and it may be misleading to assess specificity when incomplete and inaccurate measures of adverse outcomc are used. They aimed for a high specificity and to obtain this, set a very high level of resistance index for their definition of an abnormal study, resulting in only 543% of their patients having a positive test. This gave many false negative results (74% for primi- gravid proteinuric hypertension) and a poor sensitivity performance of their test. They could have emphasized the high degrees of sensitivity obtained by using the most abnormal waveform (‘worst’ RI >0.58), whilst accepting a reduction in specificity. Their result of 82% sensi- tivity for ‘worst’ RI (>0.58) in predicting proteinuiric hypertension compares favourably with that of 63% obtained by Steel et a/. (1990) who also used the most abnormal waveform obtained. By reducing the sensitivity of the test in favour of a high specificity they have devalued a potentially useful tool for the early identification of the ‘at risk’ pregnancy. Such a move could be justified if any resul- tant intervention was extreme, but in this case more intensive monitor- ing of the pregnancy would be the usual course. In this case, high sensitivity is the most clinically useful test characteristic. Roger A. Fay Clinical Lecturer David Ellwood Associate Professor Department of Obstetrics and Gynaecology University of Sydney Nepean Hospital Penrith 2750, NSW Australia References Abdella T. N., Sibai B. M., Hayes J. M. & Anderson G. D. (1984) Relationship of hypertensive disease to abruptio placentae. Ohstet Gynecol63, 365-310. Altman D. G. & Hytten F. E. (1989) Intrauterine growth retardation: lets be clear about it. Br J Ohstet Gynaecol96, I 127-1 128. Bewley S., Cooper D. & Campbell S. (1991) Doppler investigation of uteroplacental blood flow resistance in the second trimester: a screening study for pre-eclampsia and intrauterine growth retar- dation. Br J Ohstet Gynaecol98, 811-879. DeWolfe F., Brosens I. & Ranaer M. (I 980) Fetal growth retardation and the maternal arterial supply of the human placenta in the absence of sustained hypertension. Br J Ohstet Gynaecol 87, 678-685. Fay R. A., Dey P. L., Saadie C. M. J., Buhl J. A. & Gebski V. J. (1991) Ponderal Index: a better definition of the ‘at risk’ group with intra- uterine growth problems than birthweight for gestational age in term infants. Aust NZ J Obstet Gynaecol31, 17-19. Miller H. C. & Hassanein K. (1971) Diagnosis of impaired growth in newborn infants. Paediatrics 48, 5 11-522. Naeye R. L. (1989) Pregnancy hypertension, placental evidences of low uteroplacental blood flow and spontaneous premature delivery. Hum Pathol20, 441-444. Robertsons W. F., Brosens I. & Dixon G. (1967) The pathological response of the vessels of the placental bed to hypertensive preg- nancy. J Path Bact 93,581-592. Steel S. A,, Pearce J. M., McPharland P. & Chamberlain G. V. P. (1990) Early Doppler ultrasound screening in prediction of hyper- tensive disorders of pregnancy. Lancet 335, 1548-155 1. Uzan S., Uzan M., Beaufils M. & Sfar R. (1989) Compared course of 527