Continuous-wave Doppler ultrasound and decreased amniotic fluid volume in pregnant women with intact or ruptured membranes

Amelia C. Cruz, MD, Barbara H. Frentzen, RN, MSN, Kevin J. Gomez, MD, Gail Allen, RN, RDMS, and Marcia Tyson-Thomas, RN, RDMS Gainesville, Florida

The cause(s) of decreased amniotic fluid in the absence of fetal anomalies and intrauterine growth retardation is not clear. A prospective study was performed to evaluate umbilical and uterine artery Doppler velocimetric results in pregnancies complicated by decreased amniotic fluid. Three medically high­risk groups were studied: women with (1) normal fluid and intact membranes, (2) decreased fluid and intact membranes, and (3) decreased fluid and ruptured membranes. The decreased fluid/intact membr~nes group had a significantly increased incidence of abnormal uterine artery waveforms (diastolic notching or absence of end-diastolic velocity); however, uterine systolic/diastolic ratios were not significantly different. The umbilical systolic/diastolic ratios were marginally higher in the intact membranes/decreased fluid group when compared with the ruptured membranes group. This study suggests that problems with maternal blood supply to the placenta may be related to decreased amniotic fluid when membranes are intact. (AM J OasTET GYNECOL 1988;159:708-14.)

Key words: Doppler, rupture of the membranes, decreased amniotic fluid

The relationship of decreased amniotic fluid to in­trauterine growth retardation 1 " and to congenital anomalies'·' has been reported in several studies. How­ever, decreased amniotic fluid is also noted in preg­nancies in which no corresponding fetal problem is detected.6.7 The cause(s) of the decreased fluid and its clinical significance are not clear in these cases.

Currently, the underlying mechanisms responsible for production and regulation of amniotic fluid volume are only modestly understood. 8 Both maternal and fetal factors are thought to contribute to fluid regulation and disturbances in volume. The use of continuous-wave Doppler studies to analyze umbilical and uterine artery velocity waveforms offers a new, noninvasive technique to evaluate the role of uteroplacental and fetoplacental circulations in the control of amniotic fluid volume. Abnormal findings in the umbilical arteries suggest a problem in the fetus or in fetally derived placental vas­culature, whereas abnormal uterine artery study results may indicate a problem with a maternal contribution. The purpose of this study was to evaluate the fetopla-

From the Department of Obstetrics and Gynecology, University of Florida College of Medicine.

Supported in part by a grant from the Department of Health and Rehabilitative Services, State of Florida, Children's Medical Ser­vices.

Presented as Official Guest at the Fiftieth Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Palm Beach, Florida, January 10-13, 1988.

Reprint requests: Amelia C. Cruz, MD, Department of Obstetrics and Gynecology, J-294, University of Florida College of Medicine, Gainesville, FL 32610.

708

cental and uteroplacental vasculatures by Doppler ve­locimetric examination in pregnancies complicated by decreased amniotic fluid volume with either (I) intact membranes or (2) ruptured membranes, as compared with (3) intact membranes and normal fluid volume.

Material and methods This prospective study was conducted between Sep­

tember 1986 and October 1987 at Shands Hospital, University of Florida, Gainesville. Seventy pregnant women were studied: 29 had intact membranes and decreased amniotic fluid (decreased fluid group); 21 had ruptured membranes and decreased amniotic fluid (ruptured membranes group); and 20 had intact mem­branes and normal amniotic fluid volume (normal fluid group). All subjects had singleton pregnancies in the third trimester and were referred to the ultrasound laboratory for evaluation of varying medical problems. As part of the routine sonographic examination, am­niotic fluid volume was classified as normal or de­creased. Amniotic fluid was considered to be decreased when the largest vertical pocket measured <3 em. This 3 em pocket has previously been shown to be re­producible and to correlate with poor perinatal out­come.9· 10 The study was approved by the institutional review board and all patients signed an informed con­sent form. Patients with sonographically detected fetal abnormalities were excluded.

Continuous-wave Doppler ultrasound studies (An­gioscan III, Unigon) were performed on each sub­ject. A 4 MHz transducer with power output of 6.5

Volume 159 Number 3

Fig. 1. Normal and abnormal umbilical arteries. A, Normal vessel with systolic/diastolic ratio of 2.6. B, Abnormal umbil­ical artery with systolic/ diastolic ratio of 3.6.

m WI cm2 was used. Blood flow velocity waveforms cre­ated by spectral analysis of the frequency shift of the Doppler signal in the umbilical and uterine arteries were evaluated. Each patient was placed in a semire­cumbent position with a towel roll propping up her left or right hip. The umbilical artery was studied first. Three or four pictures of velocity waveforms of equal height and intensity (at least 12 cardiac cycles) were measured and averaged. A ratio of systolic height to end-diastolic trough was calculated (Fig. 1). Values <3 were considered to be normal; values ;;.:3 were taken as abnormal.n Each uterine artery was then studied in a similar manner and all systolic/diastolic ratio values were averaged. Values <2.7 were considered normal; those ;;.:2.7 were taken as abnormal.n In addition, wave­form patterns were analyzed (Fig. 2). The umbilical and uterine arteries have characteristic and easily iden­tifiable waveform patterns. Uterine arteries are char­acterized by notching in early pregnancy, but as pe­ripheral resistance decreases with advancing preg-

Doppler and decreased amniotic fluid volume 709

Fig. 2. Normal and abnormal uterine arteries. A, Normal ves­sel with systolic/diastolic notch of 2 and no diastolic notch. B, Abnormal uterine artery with prominent diastolic notch. C, Abnormal uterine artery with absence of end-diastolic velocity.

nancy, the diastolic velocity level should increase and the notching disappears by 26 weeks' gestation.'' The waveform configuration was classified as normal or abnormal.

710 Cruz et al.

Table I. Demographic and pregnancy-related factors

Decreased fluid, intact membranes (n = 29)

Age (yr) 25.1 ± 5.2* Gestational age (wk) 36.3 ± 2.6* Parity

0 II I 7

;:;.2 11 Race

Black 7 White 21 Hispanic 1

Smoking Yes 13 No 16

Hypertension I preeclampsia 5 Gestational diabetes 0 IUGR* ll Preterm contractions 6 Drug abuse I Maternal age ;:;.35 yr I

Normal fluid, intact membranes (n = 20)

25.4 ± 6.2* 34.7 ± 2.7*

6 8 6

8 11 1

6 14 5 3 4 0 I 2

September 1988 Am ] Obstet Gynecol

Ruptured membranes (n = 21) p Value

21.7 ± 3.3* NS 33.4 ± 3.5* NS

lO NS 9 2

8 NS 13 0

7 NS 14 0 NS 2 NS 5 NS 2 NS 2 NS 0 NS

*Intrauterine growth retardation suspected clinically but not confirmed by ultrasound measurements.

Table II. Umbilical systolic/diastolic ratios

Normal Abnormal

No.j % No.J% p Value

Decreased fluid, intact 22 76 membranes (n = 29)

Ruptured membranes 21 100

7 24 )

0 0.016* 0

(n = 21) Decreased fluid, intact 22 76

membranes (n = 29)

Normal fluid, intact 17 85

7 24 ) NS

I5 3 membranes (n = 20)

Ruptured membranes 21 100 0 0 (n = 21) NS

Normal fluid, intact 17 85 3 15 membranes (n = 20)

*Fisher exact test, a level (Bonferroni procedure) = 0.017.

Data were analyzed by )(2 analysis, the Fisher exact test, and analysis of variance. Differences were consid­ered significant at p ~ 0.05 for )(2 and analysis of var­iance. To compare umbilical and uterine systolic/dia­stolic ratios, three Fisher exact tests were required in­stead of one )(2 result, 'because several of the expected values were <5. Because three tests were performed, the Bonferroni procedure,12 which divides the o. level by the number of tests performed, was done. The cor­rected o. level was therefore 0.017 when comparisons were made of the umbilical and uterine systolic/dia­stolic ratios.

Table III. Uterine systolic I diastolic ratio

Normal Abnormal

No.j % No.j% p Value

Decreased fluid, intact 26 90 3 membranes (n = 29)

Ruptured membranes 21 100 0 (n = 21)

Decreased fluid, intact 26 90 membranes (n = 29)

Normal fluid, intact 20 IOO

3 10 ) O NS

()

membranes (n = 20)

Ruptured membranes 21 100 0 (n = 21)

Normal fluid, intact 20 100 0 membranes (n = 20)

Table IV. Waveform configurations

p Value

Decreased fluid, intact 11 38 18 62 0.002* membranes (n = 29)

Normal fluid, intact 17 85 3 15 membranes (n = 20)

Ruptured membranes 18 86 3 14 (n = 21)

*x2 analysis.

Volume 159 Number 3

Table V. Pregnancy outcome

Doppler and decreased amniotic fluid volume 711

Arterial cord pH Venous cord pH Ponderal index

Decreased fluid, intact membranes Ruptured membranes Normal fluid, intact membranes

7.28 ± 0.05 7.30 ± 0.08 7.26 ± 0.05

7.36 ± 0.05 7.39 ± 0.05 7.33 ± 0.05

2.46 ± 0.24 2.30 ± 0.27 2.36 ± 0.19

Analysis of variance results nonsignificant.

Table VI. Doppler results and neonatal outcome by size for gestational age

Umbilical SID ratio Uterine SID ratio Configuration

Normal Abnormal Normal Abnormal Normal Abnormal

No. I % No. j % No. I % No. I % No. I % No. I %

Decreased fluid, intact membranes AGA 21 72 5 17 23 79 3 10 9 31 17 59 SGA 0 2 8 2 8 0 1 3 1 3 LGA 1 3 0 1 3 0 1 3 0

Ruptured membranes AGA 21 100 0 21 100 0 18 86 3 14 SGA 0 0 0 0 0 0 LGA 0 0 0 0 0 0

Normal fluid, intact membranes AGA 15 75 2 10 17 85 0 15 75 2 10 SGA 0 1 5 1 5 0 0 1 5 LGA 2 10 0 2 10 0 2 10 0

SID, Systolic/ diastolic; AGA, appropriate for gestational age; SGA, small for gestational age; LGA, large for gestational age.

Results The demographic characteristics and pregnancy

complications in the three groups were comparable (Table I).

The umbilical artery systolic/ diastolic ratios were sig­nificantly higher in the group with decreased amniotic fluid and intact membranes when compared with the group with decreased amniotic fluid and ruptured membranes (p = 0.016). Seven fetuses in the decreased fluid group had abnormal umbilical systolic/diastolic ratio results. Of these seven, one was delivered preterm, two were delivered by cesarean section for fetal distress, two were small for gestational age at delivery, and two had no problems at birth. Of the 22 fetuses with normal Doppler results, one was delivered early due to mater­nal preeclampsia (umbilical systolic/ diastolic ratio was normal, but the uterine systolic/ diastolic ratio was ab­normal) and all other babies were without problems at delivery. There was, however, no difference between the decreased amniotic fluid groups (with intact or rup­tured membranes) when compared with the group with intact membranes and normal amniotic fluid (Table II).

The systolic/ diastolic ratios in the uterine arteries did not differ significantly among the three groups (Table III). Waveform configurations were then compared.

Likewise, the umbilical artery waveforms were not ab­normal in any of the three groups.

The incidence of abnormal uterine artery waveform configurations was significantly higher in the group with decreased amniotic fluid and intact membranes when compared with the two other groups. Sixty-two percent of the waveforms were abnormal in the intact membranes group, compared with 14% in the ruptured membranes group and 15% in the control group. The abnormal configurations included diastolic notching and absence of forward flow in end diastole (Table IV) that was significant at the p = 0.002 level.

Although it is difficult to know which parameters are most informative regarding neonatal outcome, we chose to analyze arterial and venous cord pH and pon­deral index, because these indices are less subjective and are standard among institutions. There were no significant differences among the three groups (Ta­ble V).

Umbilical and uterine artery systolic/ diastolic ratios and waveform configurations were evaluated for the three groups in relation to neonatal size for gestational age determined by Ballard modification of the Dubo­witz score (Table VI). There were three neonates who were small for gestational age, two in the decreased

712 Cruz et al.

fluid/intact membranes group and one in the normal fluid/intact membranes group. All had abnormal um­bilical systolic I diastolic ratios.

Comment Ultrasound evaluation of amniotic fluid in conjunc­

tion with continuous-wave Doppler studies offers the opportunity for noninvasive and repetitive studies on the role of the fetoplacental and uteroplacental circu­lations in the dynamics of amniotic fluid in human subjects.

A comprehensive review of amniotic fluid by Seeds" indicates that the increase in water that represents "approximately 30 to 40 ml per day near term ... far exceeds any intrauterine or fetal metabolic source and is most likely supplied by net transfer across the pla­centa from mother to fetus." Fetal voiding and swal­lowing have been shown to play an important role in the control of amniotic fluid volume; fetuses with renal agenesis frequently have markedly reduced fluid vol­ume," and polyhydramnios is associated with fetuses who have esophageal atresia.' 4 However, some inves­tigators have concluded that fetal voiding or swallowing rates are not always different in pregnancies with nor­mal fluid, oligohydramnios, or polyhydramnios. 15

· 16

Uterine circulatory changes may provide the clue in these cases.

In this study, abnormal waveform configurations in the uterine arteries were noted at a significantly higher incidence in the group with decreased amniotic fluid but intact membranes when compared with the normal fluid group and with the ruptured membranes group. The abnormal configurations were diastolic notching and absence of end-diastolic velocity. The physiologic mechanisms responsible for these findings are not clear. Because the abnormal findings were noted in the group with decreased amniotic fluid and intact membranes, but not in the group with decreased amniotic fluid and ruptured membranes, it can be suggested that the pres­ence of decreased amniotic fluid alone did not account for the abnormal waveform configurations.

These results suggest a role for the uteroplacental circulation in the control of amniotic fluid volume and may indicate that the maternal circulation contributes more to the amniotic fluid volume in the third trimester than previously thought. Additional studies to evaluate this hypothesis are in progress.

REFERENCES 1. Manning FA, Hill LM, Platt LD. Qualitative amniotic fluid

volume determination by ultrasound: antepartum detec­tion of intrauterine growth retardation. AM J OBSTET Gv­NECOL 1981;139:254-8.

2. Phillipson EH, Sokol RJ, Williams T. Oligohydramnios: clinical association and predictive value for intrauterine growth retardation. AM J 0BSTET GYNECOL 1983;146: 271-8.

September 1988 Am J Obstet Gynecol

3. Chamberlain PF, Manning FA, Morrison I, Harman CR, Lange IR. Ultrasound evaluation of amniotic fluid vol­ume. I. The relationship of marginal and decreased am­niotic fluid volumes to perinatal outcome. AM J OBSTET GYNECOL 1984; 150:245-9.

4. Mercer LJ, Brown LG. Fetal outcome with oligohydram­nios in the second trimester. Obstet Gynecol 1986;67: 840-2.

5. Resnik R. Amniotic fluid. In: Creasy RK, Resnik R, eds. Maternal-fetal medicine: principles and practice. Phila­delphia: WB Saunders, 1984:135-9.

6. Mercer LJ, Brown LG, Petres RE, Messer RH. A survey of pregnancies complicated by decreased amniotic fluid. AM j 0BSTET GYNECOL 1984;149:355-61.

7. Queenan JT, Thompson W, Whitfield CR, Shah SI. Am­niotic fluid volumes in normal pregnancies. AM J 0BSTET GYNECOL 1972;114:34-8.

8. Seeds AE. Current concepts of amniotic fluid dynamics. AM] 0BSTET GYNECOL 1980;138:575-86.

9. Hardt N, Johnston M, Allen G, Tyson-Thomas M, Cruz A. A classification system for amniotic fluid decrease: a useful indicator of increased perinatal risk. J Ultrasound med l985;4(suppl): 182.

10. Allen G, Tyson-Thomas M, Cruz AC. A qualitative method to classify amniotic fluid decrease performed with real time ultrasound imaging. J Ultrasound Med 1985: 4(suppl): 140.

11. Schulman H. The clinical implications of Doppler ultra­sound analysis of the uterine and umbilical series. AM J 0BSTET GYNECOL 1987;156:889-93.

12. Marascuilo LA, McSweeney M. Probabilistic and statistical concepts necessary for nonparametric and distribution­free procedures. In: Marascuilo LA, McSweeney M, eds. Nonparametric and distribution-free methods for the so­cial sciences. Pacific Grove, Calif: Brooks/Cole, 1977: 28-9.

13. Bain AD, Scott JS. Renal agenesis and severe urinary tract dysplasia. A review of 50 cases with particular ref­erence to the associated anomalies. Br Med J 1960; I: 841-5.

14. Scott JS, Wilson LK. Hydramnios as an early sign of oe­sophageal atresia. Lancet 1957;2:569.

15. Abramovich DR, Garden A, Jandial L, Page KR. Fetal swallowing and voiding in relation to hydramnios. Obstet Gynecol 1979;54: 15-20.

16. Van Otterlo LC, Wladimiroff JW, Wallenburg HCS. Re­lationship between fetal urine production and amniotic fluid volume in normal pregnancy and pregnancy com­plicated by diabetes. Br J Obstet Gynaecol 1977;84:205-9.

Editors' note: This manuscript was revised after these discussions were presented.

Discussion DR. HUGH W. RANDALL, Atlanta, Georgia (Official

Guest). It has been more than 10 years since Fitzgerald and Drumm' reported their sonographic observations on the umbilical circulation. There has been a bur­geoning body of obstetric literature in the past few years on Doppler evaluation of fetoplacental anad uteropla­cental blood flow. McCallum et al.2 noted velocity wave­form differences between normal and some compli­cated pregnancies, suggesting that impaired umbilical blood flow could be detected. Velocimetric studies by Doppler ultrasound can provide useful noninvasive in­formation to the clinician in selected high-risk patients.

In 1980 Stuart et a!.' serially calculated the peak sys-

Volume 159 Number 3

tolic to end-diastolic flow velocity ratio in 18 normal pregnancies from 16 weeks' gestation and periodically thereafter until delivery. They demonstrated a fall in placental resistance to blood flow with advancing ges­tational age, as evidenced by a decline in the ratio. This ratio is independent of the probe/vessel angle and has become the benchmark for the evaluation of arterial blood flow.

The current study by Cruz et al. has implicated the maternal circulation in the pathophysiology of de­creased amniotic fluid in a preselected group of high­risk patients. There are several areas that need to be clarified; therefore, the following questions are posed.

How were observer bias and variability minimized? This study investigates the relationship between am­niotic fluid volume and flow velocity waveforms in selected high-risk pregnancies. Both of these factors are subject to observer variability. The former was con­sidered abnormal if the largest vertical pocket mea­sured <3 em. There are limitations to using largest or maximum vertical pocket measurements to predict ab­normal fetal growth! The studies cited by Cruz et al. used various sonographic criteria to define oligohy­dramnios ranging from < 1 to <3 em. A 3 em rule may not be the best predictor for purposes of this study. In addition, intraobserver variability of vertical diameter may be high. Patterson et al. 5 have reported that the average of three diameters of the largest pocket is more reproducible than is the maximum vertical pocket.

The study shows no significant difference in uterine arterial systolic/ diastolic ratios, but does find a signif­icant difference in waveform configurations between the intact membranes group and the other two groups. These results are worrisome, as apparently no objective differences were noted in the ratios; only subjective differences in waveform interpretation were found. Again, intraobserver variability should be taken into account. In addition, the interpretation should be made by an independent observer without knowledge of pa­tient classification (intact versus ruptured versus con­trol) to avoid potential bias. Such prior knowledge could inject serious methodologic problems into a study of this type.

Were pathologic evaluations performed on the pla­centas? The uteroplacental circulation may play a role in regulation of amniotic fluid volume during the third trimester. However, to implicate the maternal circula­tion on the basis of these results seems premature. His­tologic correlation between umbilical arterial flow ve­locity waveforms and placental resistance has identified a specific microvascular lesion in the placenta.6 This increased resistance should be reflected in both um­bilical and uterine arterial flow velocity waveforms. Perhaps the differences noted here implicate a placen­tal pathologic condition instead of maternal vascular disease.

Which came first, the decreased fluid or the abnormal flow velocity waveforms? A cause-and-effect relation-

Doppler and decreased amniotic fluid volume 713

Fig. 1. Umbilical artery waveform with absent end-diastolic velocity at 19 weeks' gestation in a patient with oligohydram­nios who subsequently was diagnosed to have lupus erythe­matosus.

ship would at least require study before the develop­ment of decreased fluid. Campbell et aJ.7 analyzed flow velocity waveforms in uterine vessels (arcuate arteries) at 16 to 18 weeks' gestation in 126 pregnancies. Based on a resistance index to compare waveforms, they found that abnormal waveforms early in pregnancy predicted a subsequently abnormal outcome in 42% of cases (n = 50), with an 87% predictive value of a neg­ative test (n = 76). Certainly, a more convincing ar­gument could be made to implicate the maternal cir­culation if abnormal uterine arterial flow velocity wave­forms preceded decreased amniotic fluid volume and abnormal umbilical flow velocity waveforms. Also, there is no mention of amniotic fluid volume before rupture in the ruptured membranes group.

What was the perinatal outcome of patients with ab­normal flow velocity waveforms? There were 10 pa­tients with abnormal umbilical systolic/ diastolic ratios, 3 with abnormal uterine systolic/ diastolic ratios, and 24 (34%) with abnormal uterine waveform configurations. The latter included patients with absent end-diastolic velocity, the most extreme waveform abnormality. Ro­chelson et al.8 found abnormal perinatal outcome in 100% of patients with absent end-diastolic velocity in the fetoplacental circulation. Even abnormal sys­tolic/ diastolic ratios without absent end-diastolic veloc­ity were significant in their study. Anecdotally, a patient referred to our ultrasound laboratory with oligohy­dramnios was found to have umbilical absent end­diastolic velocity at 19 weeks' gestation (Fig. 1 ). She developed a severe preeclampsia-type picture at 22 weeks, with intrauterine fetal death; her pregnancy was terminated. She subsequently was diagnosed as having systemic lupus erythematosus. Although a com­parison of perinatal outcome between patients with normal versus abnormal flow velocity waveforms was not the thrust of the study of Cruz et al., a significant

714 Cruz et al.

outcome difference between these two groups would help to validate the interpretations.

REFERENCES l. Fitzgerald DE, Drumm JE. Noninvasive measurement of

human fetal circulation using ultrasound: a new method. Br MedJ 1977;2:1450-l.

2. McCallum WD, Williams CS, Nape! S, Daigle RE. Fetal blood velocity wave forms. AM J 0BSTET GYNECOL 1978; 132:425-9.

3. Stuart B, Drumm 1. Fitzgerald DE, Duignan NM. Fetal blood velocity wave forms in normal pregnancy. Br J Obstet Gynaecol 1980;87 :780-5.

4. Bottoms SF, Welch RA, Zador IE, Sokol RJ. Limitations of using maximum vertical pocket and other sonographic evaluations of amniotic fluid volume to predict fetal growth: technical or physiologic? AM 1 0BSTET GYNECOL 1986;155:145-8.

5. Patterson RM, Prihoda TJ, Pouliot MR. Sonographic am­niotic fluid measurement and fetal growth retardation: a reappraisal. AM J 0BSTET GYNECOL 1987:15 7: 1406-10.

6. Giles WB, Trudinger BJ, Baird PJ. Fetal umbilical artery flow velocity wave forms and placental resistance: patho­logical correlation. Br 1 Obstet Gynaecol 1985;92:31-8.

7. Campbell S, Pearce JMF, Hackett G, Cohen-Overbeek T. Hernandez C. Qualitative assessment of uteroplacental blood flow: screening test for high-risk pregnancies. Obstet Gynecol 1986;68:649-53.

8. Rochelson B, Schulman H, Farmakides G, et a!. The sig­nificance of absent end-diastolic velocity in umbilical artery velocity waveforms. AM ] 0BSTET GYNECOL 1987; 156: 1213-8.

DR. CRUZ (Closing). To answer the last question first, there were some patients with hypertension in all three groups and there was no statistically significant differ­ence between the groups. Regarding the amniotic fluid classification, I acknowledge the fact that different au­thors classify different amounts of amniotic fluid either on largest vertical pocket or average size of different pockets. At our institution, we have been using this type

September 1988 Am .J Obstet Gytwcol

of classification for the past 7 years. This classification was presented at the American Institute of Ultrasound in Medicine meeting; at that time we also showed cor­relation between this classification and a higher inci­dence of abnormal fetal outcomes.'· 2 We have basically continued with this classification because many of our studies have been predicated on this classification. Only two persons were doing amniotic fluid measurements, and the bias between their measurements has already been tested and is not felt to be an issue. In terms of the Doppler studies, again only two persons did all of the measurements, so we feel that the bias is minimal.

We also planned to do no additional studies because we really were not sure if the abnormal Doppler study findings were present before or after the decrease in amniotic fluid. We looked at the pregnancy outcome and decided not to include it in this presentation (data included in final article). If one looks at cord gases as a measure of intrauterine growth retardation, there were no significant differences among any of the three groups. Inasfar as the patients with ruptured mem­branes are concerned, not all of them underwent an ultrasound examination before membrane rupture; thus I can make no comment on the amount of amniotic fluid before membranes ruptured. What we were con­cerned with was that the findings would be related to pressure on the umbilical blood vessels, which we would then use as a comparison group.

REFERENCES I. Hardt N, Johnston M. Allen G, Tyson-Thomas M, Cruz

A. A classification system for amniotic fluid decrease: a useful indicator of increased perinatal risk. J Ultrasound Med 1985;4(suppl): 182.

2. Allen G. Tyson-Thomas M, Cruz AC. A qualitative method to classify amniotic fluid decrease with real time ultrasound imaging. 1 Ultrasound Med 1985;4(suppl):l40.