- Email: [email protected]
Does amniotic fluid index affect the accuracy of estimated fetal weight in pre term premature rupture of membranes?
Allen et al.
2. Onyskowova Z, Dolezal A, Jedlicka V. The frequency and· the character of malformations in multiple birth (a preliminary report). Teratology 1971 ;4:496-7. 3. Layde PM, Erickson ]D, Falek A, et al. Congenital malformations in twins. Am] Hum Genet 1980;32:69-78. 4. Hay S, Wehrug DA. Congenital malformations in twins. Am] Hum Genet 1970;22:662-78. 5. Myrianthopoulos NC. Congenital malformations in twins: epidemiologic survey. Birth Defects 1975;11:1-39. 6. American College of Obstetricians and Gynecologists. Multiple gestation. Washington DC: American College of Obstetricians and Gynecologists, August 1989; technical bulletin no 131. 7. American College of Obstetricians and Gynecologists. Ul-
October 1991 Am J Obstet Gynecol
trasound in pregnancy. Washington DC: American College of Obstetricians and Gynecologists, May 1988; technical bulletin no 116. 8. Holmes LB. Current concepts in genetics: congenital malformations. N Engl] Med 1976;295:204-7. 9. Gomez K], Dowdy K, Allen G, et al. Evaluation of ultrasound diagnosis of fetal anomalies in women with pregestational diabetes: University of Florida experience. AM ] OBSTET GYNECOL 1988;159:584-6. 10. Co pel ]A, Pilu G, Green], et al. Fetal echocardiographic screening of congenital heart disease: the importance of the four-chamber view. AM ] OBSTET GYNECOL 1987;157:648-55.
Does amniotic fluid index affect the accuracy of estimated fetal weight in preterm premature rupture of membranes? Julianne S. Toohey, MD, David F. Lewis, MD, James A. Harding, MD, Michael Crade, MD, Tamerou Asrat, MD, Carol A. Major, MD, Thomas J. Garite, MD, and Manuel Porto, MD Orange and Long Beach, California Estimated fetal weights playa critical role in the management scheme of patients with preterm premature rupture of membranes but are often technically difficult to obtain in these patients because of low amniotic fluid volume. Previous studies have had conflicting data as to the accuracy of estimated fetal weights in preterm premature rupture of membranes. This study was undertaken to evaluate the effect of amniotic fluid index on the accuracy of estimated fetal weights in pregnancies complicated by preterm premature rupture of membranes. Over a 2-year period at Long Beach Memorial Medical Center, 98 patients with preterm premature rupture of membranes who had an ultrasonographic examination with estimated fetal weights and amniotic fluid index performed within 48 hours of delivery were identified and compared with a control group of 55 patients in preterm labor with normal amniotic fluid index for gestational age, also obtained within 48 hours of delivery. Shepard and Hadlock formulas were used to estimate fetal weight. Results were measured in percent error from the actual birth weight. All birth weights were <2000 gm. No statistical differences were identified. The value of amniotic fluid index did not affect the accuracy of predicted estimated fetal weight in preterm premature rupture of membranes. Predicted estimated fetal weight of patients with preterm premature rupture of membranes appears to be as accurate as predicted estimated fetal weight in pregnanCies with normal amniotic fluid volumes. (AM J OasTET GVNECOL 1991 ;165:1060-2.)
Key words: Estimated fetal weight, pre term premature rupture of membranes, amniotic fluid index Ultrasonographically estimated fetal weight is often used to make critical management decisions in pre term From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of California, Irvine, UCI Medical Center, and Long Beach Memorial Women's Medical Center. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetricians, San Francisco, California, January 28-February 2, 1991. Reprint requests: Julianne S. Toohey, MD, Department of Obstetrics and Gynecology, University of California, Irvine, UCI Medical Center, 101 The City Dr., Orange, CA 92668. 616130770 1060·
premature rupture of membranes. Previous studies have had conflicting reports regarding the reliability of estimated fetal weight, leaving the obstetrician unsure of the usefulness of these data. It has been suggested that low amniotic fluid volume leads to compression of the head and abdomen, in addition to reducing the resolution of the uitrasonographic image, rendering the estimated fetal weight inaccurate and unreliable.!. 2 No previous study has evaluated the effect of accuracy of estimated fetal weights and how this correlated
Volume 165 Number 4, Part 1
Accuracy of estimated fetal weight in preterm premature membrane rupture
with amniotic fluid index in pregnancies complicated by preterm premature rupture of membranes. This study was designed to evaluate the effect of oligohydramnios on the accuracy of estimated fetal weights in pregnancies complicated by preterm premature rupture of membranes.
Material and methods Medical records of 98 patients delivered at Long Beach Memorial Women's Medical Center with the diagnosis of pre term premature rupture of membranes and birth weights of <2000 gm were retrospectively analyzed. Preterm premature rupture of membranes was defined as rupture of membranes occurring before 36 weeks. It was confirmed in all patients by sterile speculum examination with pooled fluid, ferning, and alkaline pH determinations (Nitrazine test). All ultrasonographic measurements and evaluations were performed by the perinatal staff within 48 hours of delivery. Amniotic fluid indexes were determined by the fourquadrant technique, as described by Phelan et aI.' Patients were stratified into four groups according to the degree of oligohydramnios detected. Estimated fetal weights were calculated by using Shepard et al" (biparietal diameter, abdominal circumference) and Hadlock et al. 5 (femur length, abdominal circumference) formulas. The control group, matched for gestational age and birth weight, consisted of 55 patients in preterm labor with intact membranes and normal amniotic fluid index for gestational age. Ultrasonographic examinations also were performed in the control group within 48 hours of delivery and by the same perinatal staff. Power analysis predicted an 80% chance of detecting a 30% difference in the percent error from actual birth weight in a total population of 98 patients. Data were analyzed with the Wilcoxon rank sum test with a p value <0.05 considered statistically significant. All comparisons were against a two-sided alternative hypothesis expressed as percent error ± 1 SD from actual birth weight.
Results Table I depicts the effect of amniotic fluid index measurements on the accuracy of predicted estimated fetal weight assessed as percent error of estimated fetal weight from actual birth weights, with the Shepard formula. Table II describes the same relationship with the Hadlock formula. The patients were grouped with regard to degree of oligohydramnios, with 43 noted to have an amniotic fluid index between 0.0 and 5.0. Not depicted in the table are 19 of these 43 patients who were determined to have severe oligohydramnios with an amniotic fluid index between 0.0 and 2.0. Twentynine patients had an amniotic fluid index between 5.9
1061
Table I. Estimated fetal weight with the Shepard formula Amniotic fluid index (cm)
Estimated fetal weight*
0.0-5.0 5.1-'9.9 >10 Control
43 29 26 55
9.42 8.57 10.12 9.65
± 7.2 ± 6.7
± 8.5 ± 8.5
*Percent error ± SD.
Table II. Estimated fetal weight with the Hadlock formula Estimated fetal weight*
Amniotic fluid index (cm)
0.0-5.0 5.1-9.9 >10 Control
43 29 26 55
9.15 9.87 12.18 10.88
± 7.2 ± 6.9 ± 10.7 ± 8.8
*Percent error ± SD.
Table III. Correlation of birth weight and estimated fetal weight in pre term premature rupture of membranes Birth weight (Ifrrt)
500-1000 1001-1500 1501-2000
No.
Shepard*
Hadlock*
35 27
9.65 ± 8.5 9.27 ± 6.7 9.40 ± 6.6
11.15 ± 9.9 10.74 ± 8.2 9.47 ± 6.6
36
*Percent error ± SD.
and 9.9, and 26 patients were determined to have fluid indexes of > 10. No statistical difference was noted in percent error between the study patients with preterm premature rupture of membranes and the control group with intact membranes, or between the subgroups of amniotic fluid volume. In addition, there were no statistical differences in the percent error calculated by either formula. Patients also were stratified with regard to birth weight, with 35 infants weighing between 500 and 1000 gm. As depicted in Table III, both formulas appear to be equivalent in their ability to predict estimated fetal weight with varying birth weights, even when addressing the very-low-birth-weight groups.
Comment Estimation of fetal weight plays a critical role in the decision-making and management scheme of patients with preterm premature rupture of membranes. However, the literature addressing the issue of accuracy of estimated fetal weight in the face of oligohydramnios remains confusing and sparse. All the studies examining estimated fetal weight in patients with preterm
1062 Toohey et al.
premature rupture of membranes published to date lack a concise and quantifiable definition of oligohydramnios and "low fluid volume," rendering the data difficult to interpret. In 1985, O'Keefe et aLI were the first authors to report that biparietal diameter measurements in patients with preterm premature rupture of membranes were inaccurate in 45 % of their 100 patients as assessed by an abnormal cephalic index and concluded that biparietal diameter measurements were unreliable in estimating fetal weight in patients with preterm premature rupture of membranes. This study did not quantify or justify the severity of oligohydramnios that resulted in distortion of the contour of the head. In 1986 a similar study looking at the effects of oligohydramnios on estimation of fetal weight by Ott et aF concluded that the cephalic index was not appreciably distorted by oligohydramnios. Again, that study did not define the degree of oligohydramnios observed in study patients. Subsequently, Bottoms et al. 6 reported an estimation of fetal weight in 26 pregnancies complicated by premature rupture of the membranes. Although all measurements were smaller in patients with premature rupture of the membrane than in control patients, the authors attributed this finding to compromised intrauterine growth rather than artifacts from physical compression caused by oligohydramnios. Benacerraf et aI.' in 1988 looked at the accuracy of ultrasonographically estimated fetal weight in a large series of 1301 patients with varying amounts of amniotic fluid volume and noted that the presence of oligohydramnios made no difference in the percent errors. In this study 17% were noted subjectively to have oligohydramnios and only 20 of 1301 fetuses had birth weights <2000 gm. Townsend et al. B in 1988 looked specifically at ultrasonographically estimated fetal weight in 53 very-Iowbirth-weight infants « 1000 gm). Twenty-four fetuses were noted to have "decreased" amniotic fluid volume subjectively. The authors reported that no statistical differences in accuracy of weight prediction in preterm deliveries were observed when patients with normal amniotic fluid were compared with those with oligohydramnios. Finally, in 1990 Valea et al. 9 compared 86 patients with preterm premature rupture of membranes at 536 weeks with 112 control patients with intact membranes matched for gestational age and were able to demonstrate that amniotic fluid volume was not an important variable in the accuracy of ultrasono-
October 1991 Am J Obstet Gynecol
graphically estimated fetal weights. However, in this study there was no mention of the percentage of patients with oligohydramnios, and the amniotic fluid was not quantified. In our study we quantified the degree of oligohydramnios and demonstrated in support of the cited studies that low amniotic volumes do not appreciably affect the accuracy of ultrasonographically estimated fetal weights in patients with preterm premature rupture of membranes. Nineteen of our study patients had severe oligohydramnios with an amniotic fluid index of oto 2 cm. In spite of this severe lack of fluid, the percent error from actual birth weight was only 7.45 ± 7.2 by the Shepard formula and 9.15 ± 7.2 by the Hadlock formula. Predicted estimated fetal weights in preterm gestations with preterm premature rupture of membranes appear to be as accurate as those predicted in patients with normal amniotic fluid volume, even in very-Iow-birth-weight infants. We conclude that, in the hands of experienced ultrasonographers, predicted estimated fetal weights may be used with confidence in the management of pregnant patients with oligohydramnios. REFERENCES 1. O'Keefe DF, Garite T], Elliott]P, Burnes PE. The accuracy of estimated gestational age based on ultrasound measurement of biparietal diameter in preterm premature rupture of the membranes. AM] OBSTET GYNECOL 1985; 151 :30912. 2. Ott W], Doyle S, Flamm A. Accurate ultrasonic estimation of fetal weight. Effect of head shape, growth patterns, and amniotic fluid volume. Am] PerinatoI1986;3:193-7. 3. Phelan]P, Smith CV, Broussard P, Small M. Amniotic fluid assessment with the four quadrant technique at 36-42 weeks gestation.] Reprod Med 1987;32:540-2. 4. Shepard M], Richards VA, Berkowitz RL, WarsofSL, Hobbins] C. An evaluation of two equations for predicting fetal weight by ultrasound. AM] OBSTET GYNECOL 1982;142:4754. 5. Hadlock FP, Harrist RB, Carpenter R], Russell LD, Seung KP. Sonographic estimation of fetal weight. Radiology 1984;150:535-42. 6. Bottoms SF, Welch RA, Zador IE, Sokol RJ. Clinical interpretation of ultrasound measurements in preterm pregnancies with premature rupture of the membranes. Obstet Gynecol 1987;69:358-62. 7. Benacerraf BR, Gelman R, Frigoletto FD. Sonographically estimated fetal weights. Accuracy and limitation. AM] OBSTET GYNECOL 1988;159:1118-21. 8. Townsend RR, Filly RA, Callen PW, Laros RK. Factors affecting prenatal sonographic estimation of weight in extremely low birth weight infants. ] Ultrasound Med 1988;7: 183-7. 9. Valea FA, Watson W], Seeds ]W. Accuracy of ultrasonic weight prediction in the fetus with preterm premature rupture of membranes. Obstet Gynecol 1990; 75: 183-4.
2. Onyskowova Z, Dolezal A, Jedlicka V. The frequency and· the character of malformations in multiple birth (a preliminary report). Teratology 1971 ;4:496-7. 3. Layde PM, Erickson ]D, Falek A, et al. Congenital malformations in twins. Am] Hum Genet 1980;32:69-78. 4. Hay S, Wehrug DA. Congenital malformations in twins. Am] Hum Genet 1970;22:662-78. 5. Myrianthopoulos NC. Congenital malformations in twins: epidemiologic survey. Birth Defects 1975;11:1-39. 6. American College of Obstetricians and Gynecologists. Multiple gestation. Washington DC: American College of Obstetricians and Gynecologists, August 1989; technical bulletin no 131. 7. American College of Obstetricians and Gynecologists. Ul-
October 1991 Am J Obstet Gynecol
trasound in pregnancy. Washington DC: American College of Obstetricians and Gynecologists, May 1988; technical bulletin no 116. 8. Holmes LB. Current concepts in genetics: congenital malformations. N Engl] Med 1976;295:204-7. 9. Gomez K], Dowdy K, Allen G, et al. Evaluation of ultrasound diagnosis of fetal anomalies in women with pregestational diabetes: University of Florida experience. AM ] OBSTET GYNECOL 1988;159:584-6. 10. Co pel ]A, Pilu G, Green], et al. Fetal echocardiographic screening of congenital heart disease: the importance of the four-chamber view. AM ] OBSTET GYNECOL 1987;157:648-55.
Does amniotic fluid index affect the accuracy of estimated fetal weight in preterm premature rupture of membranes? Julianne S. Toohey, MD, David F. Lewis, MD, James A. Harding, MD, Michael Crade, MD, Tamerou Asrat, MD, Carol A. Major, MD, Thomas J. Garite, MD, and Manuel Porto, MD Orange and Long Beach, California Estimated fetal weights playa critical role in the management scheme of patients with preterm premature rupture of membranes but are often technically difficult to obtain in these patients because of low amniotic fluid volume. Previous studies have had conflicting data as to the accuracy of estimated fetal weights in preterm premature rupture of membranes. This study was undertaken to evaluate the effect of amniotic fluid index on the accuracy of estimated fetal weights in pregnancies complicated by preterm premature rupture of membranes. Over a 2-year period at Long Beach Memorial Medical Center, 98 patients with preterm premature rupture of membranes who had an ultrasonographic examination with estimated fetal weights and amniotic fluid index performed within 48 hours of delivery were identified and compared with a control group of 55 patients in preterm labor with normal amniotic fluid index for gestational age, also obtained within 48 hours of delivery. Shepard and Hadlock formulas were used to estimate fetal weight. Results were measured in percent error from the actual birth weight. All birth weights were <2000 gm. No statistical differences were identified. The value of amniotic fluid index did not affect the accuracy of predicted estimated fetal weight in preterm premature rupture of membranes. Predicted estimated fetal weight of patients with preterm premature rupture of membranes appears to be as accurate as predicted estimated fetal weight in pregnanCies with normal amniotic fluid volumes. (AM J OasTET GVNECOL 1991 ;165:1060-2.)
Key words: Estimated fetal weight, pre term premature rupture of membranes, amniotic fluid index Ultrasonographically estimated fetal weight is often used to make critical management decisions in pre term From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of California, Irvine, UCI Medical Center, and Long Beach Memorial Women's Medical Center. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetricians, San Francisco, California, January 28-February 2, 1991. Reprint requests: Julianne S. Toohey, MD, Department of Obstetrics and Gynecology, University of California, Irvine, UCI Medical Center, 101 The City Dr., Orange, CA 92668. 616130770 1060·
premature rupture of membranes. Previous studies have had conflicting reports regarding the reliability of estimated fetal weight, leaving the obstetrician unsure of the usefulness of these data. It has been suggested that low amniotic fluid volume leads to compression of the head and abdomen, in addition to reducing the resolution of the uitrasonographic image, rendering the estimated fetal weight inaccurate and unreliable.!. 2 No previous study has evaluated the effect of accuracy of estimated fetal weights and how this correlated
Volume 165 Number 4, Part 1
Accuracy of estimated fetal weight in preterm premature membrane rupture
with amniotic fluid index in pregnancies complicated by preterm premature rupture of membranes. This study was designed to evaluate the effect of oligohydramnios on the accuracy of estimated fetal weights in pregnancies complicated by preterm premature rupture of membranes.
Material and methods Medical records of 98 patients delivered at Long Beach Memorial Women's Medical Center with the diagnosis of pre term premature rupture of membranes and birth weights of <2000 gm were retrospectively analyzed. Preterm premature rupture of membranes was defined as rupture of membranes occurring before 36 weeks. It was confirmed in all patients by sterile speculum examination with pooled fluid, ferning, and alkaline pH determinations (Nitrazine test). All ultrasonographic measurements and evaluations were performed by the perinatal staff within 48 hours of delivery. Amniotic fluid indexes were determined by the fourquadrant technique, as described by Phelan et aI.' Patients were stratified into four groups according to the degree of oligohydramnios detected. Estimated fetal weights were calculated by using Shepard et al" (biparietal diameter, abdominal circumference) and Hadlock et al. 5 (femur length, abdominal circumference) formulas. The control group, matched for gestational age and birth weight, consisted of 55 patients in preterm labor with intact membranes and normal amniotic fluid index for gestational age. Ultrasonographic examinations also were performed in the control group within 48 hours of delivery and by the same perinatal staff. Power analysis predicted an 80% chance of detecting a 30% difference in the percent error from actual birth weight in a total population of 98 patients. Data were analyzed with the Wilcoxon rank sum test with a p value <0.05 considered statistically significant. All comparisons were against a two-sided alternative hypothesis expressed as percent error ± 1 SD from actual birth weight.
Results Table I depicts the effect of amniotic fluid index measurements on the accuracy of predicted estimated fetal weight assessed as percent error of estimated fetal weight from actual birth weights, with the Shepard formula. Table II describes the same relationship with the Hadlock formula. The patients were grouped with regard to degree of oligohydramnios, with 43 noted to have an amniotic fluid index between 0.0 and 5.0. Not depicted in the table are 19 of these 43 patients who were determined to have severe oligohydramnios with an amniotic fluid index between 0.0 and 2.0. Twentynine patients had an amniotic fluid index between 5.9
1061
Table I. Estimated fetal weight with the Shepard formula Amniotic fluid index (cm)
Estimated fetal weight*
0.0-5.0 5.1-'9.9 >10 Control
43 29 26 55
9.42 8.57 10.12 9.65
± 7.2 ± 6.7
± 8.5 ± 8.5
*Percent error ± SD.
Table II. Estimated fetal weight with the Hadlock formula Estimated fetal weight*
Amniotic fluid index (cm)
0.0-5.0 5.1-9.9 >10 Control
43 29 26 55
9.15 9.87 12.18 10.88
± 7.2 ± 6.9 ± 10.7 ± 8.8
*Percent error ± SD.
Table III. Correlation of birth weight and estimated fetal weight in pre term premature rupture of membranes Birth weight (Ifrrt)
500-1000 1001-1500 1501-2000
No.
Shepard*
Hadlock*
35 27
9.65 ± 8.5 9.27 ± 6.7 9.40 ± 6.6
11.15 ± 9.9 10.74 ± 8.2 9.47 ± 6.6
36
*Percent error ± SD.
and 9.9, and 26 patients were determined to have fluid indexes of > 10. No statistical difference was noted in percent error between the study patients with preterm premature rupture of membranes and the control group with intact membranes, or between the subgroups of amniotic fluid volume. In addition, there were no statistical differences in the percent error calculated by either formula. Patients also were stratified with regard to birth weight, with 35 infants weighing between 500 and 1000 gm. As depicted in Table III, both formulas appear to be equivalent in their ability to predict estimated fetal weight with varying birth weights, even when addressing the very-low-birth-weight groups.
Comment Estimation of fetal weight plays a critical role in the decision-making and management scheme of patients with preterm premature rupture of membranes. However, the literature addressing the issue of accuracy of estimated fetal weight in the face of oligohydramnios remains confusing and sparse. All the studies examining estimated fetal weight in patients with preterm
1062 Toohey et al.
premature rupture of membranes published to date lack a concise and quantifiable definition of oligohydramnios and "low fluid volume," rendering the data difficult to interpret. In 1985, O'Keefe et aLI were the first authors to report that biparietal diameter measurements in patients with preterm premature rupture of membranes were inaccurate in 45 % of their 100 patients as assessed by an abnormal cephalic index and concluded that biparietal diameter measurements were unreliable in estimating fetal weight in patients with preterm premature rupture of membranes. This study did not quantify or justify the severity of oligohydramnios that resulted in distortion of the contour of the head. In 1986 a similar study looking at the effects of oligohydramnios on estimation of fetal weight by Ott et aF concluded that the cephalic index was not appreciably distorted by oligohydramnios. Again, that study did not define the degree of oligohydramnios observed in study patients. Subsequently, Bottoms et al. 6 reported an estimation of fetal weight in 26 pregnancies complicated by premature rupture of the membranes. Although all measurements were smaller in patients with premature rupture of the membrane than in control patients, the authors attributed this finding to compromised intrauterine growth rather than artifacts from physical compression caused by oligohydramnios. Benacerraf et aI.' in 1988 looked at the accuracy of ultrasonographically estimated fetal weight in a large series of 1301 patients with varying amounts of amniotic fluid volume and noted that the presence of oligohydramnios made no difference in the percent errors. In this study 17% were noted subjectively to have oligohydramnios and only 20 of 1301 fetuses had birth weights <2000 gm. Townsend et al. B in 1988 looked specifically at ultrasonographically estimated fetal weight in 53 very-Iowbirth-weight infants « 1000 gm). Twenty-four fetuses were noted to have "decreased" amniotic fluid volume subjectively. The authors reported that no statistical differences in accuracy of weight prediction in preterm deliveries were observed when patients with normal amniotic fluid were compared with those with oligohydramnios. Finally, in 1990 Valea et al. 9 compared 86 patients with preterm premature rupture of membranes at 536 weeks with 112 control patients with intact membranes matched for gestational age and were able to demonstrate that amniotic fluid volume was not an important variable in the accuracy of ultrasono-
October 1991 Am J Obstet Gynecol
graphically estimated fetal weights. However, in this study there was no mention of the percentage of patients with oligohydramnios, and the amniotic fluid was not quantified. In our study we quantified the degree of oligohydramnios and demonstrated in support of the cited studies that low amniotic volumes do not appreciably affect the accuracy of ultrasonographically estimated fetal weights in patients with preterm premature rupture of membranes. Nineteen of our study patients had severe oligohydramnios with an amniotic fluid index of oto 2 cm. In spite of this severe lack of fluid, the percent error from actual birth weight was only 7.45 ± 7.2 by the Shepard formula and 9.15 ± 7.2 by the Hadlock formula. Predicted estimated fetal weights in preterm gestations with preterm premature rupture of membranes appear to be as accurate as those predicted in patients with normal amniotic fluid volume, even in very-Iow-birth-weight infants. We conclude that, in the hands of experienced ultrasonographers, predicted estimated fetal weights may be used with confidence in the management of pregnant patients with oligohydramnios. REFERENCES 1. O'Keefe DF, Garite T], Elliott]P, Burnes PE. The accuracy of estimated gestational age based on ultrasound measurement of biparietal diameter in preterm premature rupture of the membranes. AM] OBSTET GYNECOL 1985; 151 :30912. 2. Ott W], Doyle S, Flamm A. Accurate ultrasonic estimation of fetal weight. Effect of head shape, growth patterns, and amniotic fluid volume. Am] PerinatoI1986;3:193-7. 3. Phelan]P, Smith CV, Broussard P, Small M. Amniotic fluid assessment with the four quadrant technique at 36-42 weeks gestation.] Reprod Med 1987;32:540-2. 4. Shepard M], Richards VA, Berkowitz RL, WarsofSL, Hobbins] C. An evaluation of two equations for predicting fetal weight by ultrasound. AM] OBSTET GYNECOL 1982;142:4754. 5. Hadlock FP, Harrist RB, Carpenter R], Russell LD, Seung KP. Sonographic estimation of fetal weight. Radiology 1984;150:535-42. 6. Bottoms SF, Welch RA, Zador IE, Sokol RJ. Clinical interpretation of ultrasound measurements in preterm pregnancies with premature rupture of the membranes. Obstet Gynecol 1987;69:358-62. 7. Benacerraf BR, Gelman R, Frigoletto FD. Sonographically estimated fetal weights. Accuracy and limitation. AM] OBSTET GYNECOL 1988;159:1118-21. 8. Townsend RR, Filly RA, Callen PW, Laros RK. Factors affecting prenatal sonographic estimation of weight in extremely low birth weight infants. ] Ultrasound Med 1988;7: 183-7. 9. Valea FA, Watson W], Seeds ]W. Accuracy of ultrasonic weight prediction in the fetus with preterm premature rupture of membranes. Obstet Gynecol 1990; 75: 183-4.