Saturday 28 November
No 8831
1992
ORIGINAL ARTICLES
Randomised controlled trial of doppler ultrasound screening of placental perfusion during pregnancy
We have done a randomised controlled trial to assess the effect on primary management and outcome of routine doppler ultrasound examinations of the umbilical and uterine arteries during pregnancy. Over 9 months, 2600 women with singleton pregnancies were recruited from a general obstetric population. Of 2475 women who delivered in hospital after 20 weeks’ gestation, 1246 had been allocated at random to receive standard antenatal care with routine doppler examinations. The first doppler ultrasound was done at 19-22 weeks’ gestation, and thereafter examinations were monthly if the pregnancy was considered high risk (192) or once at 32 weeks if considered low risk (1054). The control group of 1229 women received standard, antenatal care without doppler ultrasonography. The study groups did not differ in number of antenatal admissions or cardiotocographs, gestational age at delivery, method of delivery, frequency of deliveries with fetal distress, or need for resuscitation or admission to the neonatal intensive care unit. More perinatal deaths occurred in the doppler group (17 vs 7, relative risk 2·4, 95% Cl 1·00-5·76), but only 1 of 11 normally formed stillbirths and none of the 4 normally formed neonatal deaths after 24 weeks’ gestation had an abnormal umbilical-artery doppler examination. We did not demonstrate any improvement in neonatal outcome by routine doppler ultrasound screening of a general obstetric population.
Introduction In women at high risk of an abnormal pregnancy, doppler ultrasonography of the umbilical artery can detect fetal compromise. Five randomised controlled studies have evaluated doppler ultrasound examinations in such pregnancies. In two studies the control group did not have doppler examinations,2,3and in another two studies the results from the control group were concealed. 1,4 The fifth study compared selective use of doppler ultrasound with routine use as part of a modified biophysical profile.5 Meta-analysis of the first four studies showed that availability of the results of umbilical-artery doppler ultrasound examinations was associated with a decrease in
perinatal mortality However, there was and perinatal
normally formed infants. significant effect on other obstetric
among no
outcome
measures.6
Screening studies7,8 indicate that there may be little value in routine doppler ultrasound examinations of all pregnancies because of poor prediction of adverse outcome. However, uterine-artery examinations done between 20 and 24 weeks of gestation have some value in predicting pregnancy-induced hypertension.9-11 Randomised controlled studies to test the value of routine doppler ultrasonography in a general obstetric population have not been done. We have carried out such a study according to a ADDRESSES. Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Queen Charlotte’s and Chelsea Hospital, London (J. A. Davies, MRCOG, J. A. D. Spencer, MRCOG); and Clinical Operational Research Unit, Department of Statistical Science, University College London, London, UK (S. Gallivan, PhD). Correspondence to Mr J. A. D. Spencer, Department of Obstetrics and Gynaecology, University College and Middlesex School of Medicine, University College London, 89-96 Chenies Mews, London WC1 E 6HX, UK
1300
TABLE I-DEMOGRAPHIC DETAILS OF STUDY GROUPS
risk-related protocol but without a protocol for management of abnormal results.
Subjects and methods Recruitment and allocation Women were eligible for inclusion if their first (booking) visit to the antenatal clinic at Queen Charlotte’s and Chelsea Hospital was before 20 weeks’ gestation. All women were sent information about the study with their initial appointment letter. Further information on the study was available from staff in the clinic and written consent was obtained from women at the time of their first attendance. At the time of their anomaly scan (19-22 weeks’ gestation), women with singleton pregnancies were allocated at random to either the doppler or control groups by cards in sealed opaque envelopes. Randomisation was done in four blocks of 500 and three blocks of 200. The study was approved by the local ethics committee. All women had standard antenatal care, mainly shared care with local general practitioners, according to hospital policy. In addition, the doppler group were given appointments for routine umbilicalartery and uterine-artery doppler ultrasound examinations according to their risk of having a small-for-gestational-age, or otherwise compromised, infant. Women with low-risk pregnancies were examined once more at 32 weeks’ gestation when they returned for a routine antenatal clinic visit. Those considered at high risk had a doppler ultrasound examination every month. Women with high-risk pregnancies were identified at the first antenatal visit (before entry into the study) by the presence of one or more of the following factors: a preexisting medical condition such
hypertension or diabetes, a previous small-for-gestational-age a previous stillbirth or neonatal death, hypertension (blood pressure > 140/90 mm Hg) in a previous pregnancy or at booking, or smoking more than 10 cigarettes a day. Any woman with a low-risk pregnancy who had an abnormal doppler ultrasound examination was managed subsequently according to the high-risk protocol. If a subsequent examination was normal, the woman was transferred back to the low-risk group. Women in the doppler group were eligible for ultrasound examinations at times other than the planned routine screenings if requested by a clinician, but it was intended that women in the control group should not have doppler ultrasound examinations at any time during pregnancy. as
livebirth,
Doppler ultrasound examinations Ultrasound examinations were done with either Vasoflo-3 or Vasoflo-4 (Oxford Soncaid) continuous-wave doppler machines.l2 Women were usually examined in a semi-recumbent position. Doppler indices were automatically calculated by the machine, which was set to give a mean value derived from six consecutive waveforms when a visually acceptable signal was frozen on the monitor screen.13 Examinations were done without real-time ultrasound imaging. Umbilical-artery waveforms were identified when the continuous negative venous signal was simultaneously visualised, and were usually obtained from the side of the maternal abdomen opposite the fetal back. Uterine-artery waveforms were identified as positive frequencies (blood flow towards the probe) along the lower lateral border of the uterus with the probe directed downwards and medially.’Z Although every attempt was made to obtain recordings during fetal apnoea and quiescence this was not
always practicable. However,
measurements were
Figures are number z unless indicated otherwise. were no significant differences between groups
There
from both uterine arteries if possible, but the result was reported as normal if only one side was abnormal since we believed that the blood supply on the placental (lower resistance) side would be the most important for pregnancy outcome. The first doppler examination was usually done by the ultrasonographer at the time of the anomaly scan. Subsequent examinations on all women were usually done by one of us (J. A. D.). We have shown previously that interoperator variability for umbilical doppler ultrasound measurements is negligible.le The lowest resistance index (RI) from the uterine arteries and the pulsatility index (PI) from the umbilical artery were reported as normal, "raised" (indices above the 95th percentile for gestation), or showing absent end-diastolic frequencies. If abnormal values were obtained for the uterine circulation at the initial examination, the test was repeated at 24 weeks before being deemed abnormal.’" All abnormal results after 24 weeks were also reported verbally to a clinician and the woman was referred to the antenatal clinic. Subsequent management was according to the clinicians concerned with the case.
Data collection and analysis Outcome data were collected from women’s notes by J. A. D. and medical coding clerk after delivery. Data were transferred to computer and validated. Data included the number of days of antenatal admission, number of cardiotocograph recordings and ultrasound scans during pregnancy, gestational age at delivery, method of delivery, birthweight, Apgar scores, need for a
TABLE ))—ANTENATAL FINDINGS IN STUDY GROUPS AND RELATIVE RISKS
only accepted
when at least six regular, consecutive, waveforms had been obtained.14 To maintain an appointment schedule, a maximum of 10 min was allocated for each routine screening. The reference ranges used for the study were considered to be the best available at the time. They had been reported recently and were based on longitudinal measurements from a nearby London population. is Although these reference ranges were constructed from arcuate artery data obtained with pulsed doppler ultrasonography, there is close agreement with other reference ranges whether produced from longitudinal or cross-sectional data on pulsed or continuous-wave machines.16,17 Values were obtained
Figures are number (%) unless indicated otherwise. *p
No 8831
1992
ORIGINAL ARTICLES
Randomised controlled trial of doppler ultrasound screening of placental perfusion during pregnancy
We have done a randomised controlled trial to assess the effect on primary management and outcome of routine doppler ultrasound examinations of the umbilical and uterine arteries during pregnancy. Over 9 months, 2600 women with singleton pregnancies were recruited from a general obstetric population. Of 2475 women who delivered in hospital after 20 weeks’ gestation, 1246 had been allocated at random to receive standard antenatal care with routine doppler examinations. The first doppler ultrasound was done at 19-22 weeks’ gestation, and thereafter examinations were monthly if the pregnancy was considered high risk (192) or once at 32 weeks if considered low risk (1054). The control group of 1229 women received standard, antenatal care without doppler ultrasonography. The study groups did not differ in number of antenatal admissions or cardiotocographs, gestational age at delivery, method of delivery, frequency of deliveries with fetal distress, or need for resuscitation or admission to the neonatal intensive care unit. More perinatal deaths occurred in the doppler group (17 vs 7, relative risk 2·4, 95% Cl 1·00-5·76), but only 1 of 11 normally formed stillbirths and none of the 4 normally formed neonatal deaths after 24 weeks’ gestation had an abnormal umbilical-artery doppler examination. We did not demonstrate any improvement in neonatal outcome by routine doppler ultrasound screening of a general obstetric population.
Introduction In women at high risk of an abnormal pregnancy, doppler ultrasonography of the umbilical artery can detect fetal compromise. Five randomised controlled studies have evaluated doppler ultrasound examinations in such pregnancies. In two studies the control group did not have doppler examinations,2,3and in another two studies the results from the control group were concealed. 1,4 The fifth study compared selective use of doppler ultrasound with routine use as part of a modified biophysical profile.5 Meta-analysis of the first four studies showed that availability of the results of umbilical-artery doppler ultrasound examinations was associated with a decrease in
perinatal mortality However, there was and perinatal
normally formed infants. significant effect on other obstetric
among no
outcome
measures.6
Screening studies7,8 indicate that there may be little value in routine doppler ultrasound examinations of all pregnancies because of poor prediction of adverse outcome. However, uterine-artery examinations done between 20 and 24 weeks of gestation have some value in predicting pregnancy-induced hypertension.9-11 Randomised controlled studies to test the value of routine doppler ultrasonography in a general obstetric population have not been done. We have carried out such a study according to a ADDRESSES. Institute of Obstetrics and Gynaecology, Royal Postgraduate Medical School, Queen Charlotte’s and Chelsea Hospital, London (J. A. Davies, MRCOG, J. A. D. Spencer, MRCOG); and Clinical Operational Research Unit, Department of Statistical Science, University College London, London, UK (S. Gallivan, PhD). Correspondence to Mr J. A. D. Spencer, Department of Obstetrics and Gynaecology, University College and Middlesex School of Medicine, University College London, 89-96 Chenies Mews, London WC1 E 6HX, UK
1300
TABLE I-DEMOGRAPHIC DETAILS OF STUDY GROUPS
risk-related protocol but without a protocol for management of abnormal results.
Subjects and methods Recruitment and allocation Women were eligible for inclusion if their first (booking) visit to the antenatal clinic at Queen Charlotte’s and Chelsea Hospital was before 20 weeks’ gestation. All women were sent information about the study with their initial appointment letter. Further information on the study was available from staff in the clinic and written consent was obtained from women at the time of their first attendance. At the time of their anomaly scan (19-22 weeks’ gestation), women with singleton pregnancies were allocated at random to either the doppler or control groups by cards in sealed opaque envelopes. Randomisation was done in four blocks of 500 and three blocks of 200. The study was approved by the local ethics committee. All women had standard antenatal care, mainly shared care with local general practitioners, according to hospital policy. In addition, the doppler group were given appointments for routine umbilicalartery and uterine-artery doppler ultrasound examinations according to their risk of having a small-for-gestational-age, or otherwise compromised, infant. Women with low-risk pregnancies were examined once more at 32 weeks’ gestation when they returned for a routine antenatal clinic visit. Those considered at high risk had a doppler ultrasound examination every month. Women with high-risk pregnancies were identified at the first antenatal visit (before entry into the study) by the presence of one or more of the following factors: a preexisting medical condition such
hypertension or diabetes, a previous small-for-gestational-age a previous stillbirth or neonatal death, hypertension (blood pressure > 140/90 mm Hg) in a previous pregnancy or at booking, or smoking more than 10 cigarettes a day. Any woman with a low-risk pregnancy who had an abnormal doppler ultrasound examination was managed subsequently according to the high-risk protocol. If a subsequent examination was normal, the woman was transferred back to the low-risk group. Women in the doppler group were eligible for ultrasound examinations at times other than the planned routine screenings if requested by a clinician, but it was intended that women in the control group should not have doppler ultrasound examinations at any time during pregnancy. as
livebirth,
Doppler ultrasound examinations Ultrasound examinations were done with either Vasoflo-3 or Vasoflo-4 (Oxford Soncaid) continuous-wave doppler machines.l2 Women were usually examined in a semi-recumbent position. Doppler indices were automatically calculated by the machine, which was set to give a mean value derived from six consecutive waveforms when a visually acceptable signal was frozen on the monitor screen.13 Examinations were done without real-time ultrasound imaging. Umbilical-artery waveforms were identified when the continuous negative venous signal was simultaneously visualised, and were usually obtained from the side of the maternal abdomen opposite the fetal back. Uterine-artery waveforms were identified as positive frequencies (blood flow towards the probe) along the lower lateral border of the uterus with the probe directed downwards and medially.’Z Although every attempt was made to obtain recordings during fetal apnoea and quiescence this was not
always practicable. However,
measurements were
Figures are number z unless indicated otherwise. were no significant differences between groups
There
from both uterine arteries if possible, but the result was reported as normal if only one side was abnormal since we believed that the blood supply on the placental (lower resistance) side would be the most important for pregnancy outcome. The first doppler examination was usually done by the ultrasonographer at the time of the anomaly scan. Subsequent examinations on all women were usually done by one of us (J. A. D.). We have shown previously that interoperator variability for umbilical doppler ultrasound measurements is negligible.le The lowest resistance index (RI) from the uterine arteries and the pulsatility index (PI) from the umbilical artery were reported as normal, "raised" (indices above the 95th percentile for gestation), or showing absent end-diastolic frequencies. If abnormal values were obtained for the uterine circulation at the initial examination, the test was repeated at 24 weeks before being deemed abnormal.’" All abnormal results after 24 weeks were also reported verbally to a clinician and the woman was referred to the antenatal clinic. Subsequent management was according to the clinicians concerned with the case.
Data collection and analysis Outcome data were collected from women’s notes by J. A. D. and medical coding clerk after delivery. Data were transferred to computer and validated. Data included the number of days of antenatal admission, number of cardiotocograph recordings and ultrasound scans during pregnancy, gestational age at delivery, method of delivery, birthweight, Apgar scores, need for a
TABLE ))—ANTENATAL FINDINGS IN STUDY GROUPS AND RELATIVE RISKS
only accepted
when at least six regular, consecutive, waveforms had been obtained.14 To maintain an appointment schedule, a maximum of 10 min was allocated for each routine screening. The reference ranges used for the study were considered to be the best available at the time. They had been reported recently and were based on longitudinal measurements from a nearby London population. is Although these reference ranges were constructed from arcuate artery data obtained with pulsed doppler ultrasonography, there is close agreement with other reference ranges whether produced from longitudinal or cross-sectional data on pulsed or continuous-wave machines.16,17 Values were obtained
Figures are number (%) unless indicated otherwise. *p
