S130
SPO Abstracts
January 1997 A m J Obstet Gynecol
439
LONGITUDINAL ASSESSMENT OF AMNIOTIC FLUID INDEX IN DIABETIC PREGNANCIES. F. Arredondo~, M. Giannotti, L. Mann, 0. Langer, Dept. of OB/GYN, UTHSC, San Antonio, TX. OBJECTIVE: To determine if threshold of amniotic fluid index (AFI) in non-diabetic wonmn can be used in diabetic pregnancies and to assess the association between the quantity of AFI and the quality of diabetic control. STUDY DESIGN: The four quadrant amniotic fluid index was measured prospectively in a longitudinal study of 170 diabetic singleton pregnancies between 15 and 41 weeks of gestation. The AFI results were stratified by week of gestation. The association between glucose profile (pre meal, post meals, mean average, HbAIC) and AFI were compared in normal and diabetic individuals. RESULTS: The AFI increased progressively from week 15 to peak at 32 weeks in normal and diabetic pregnancies. The AFI in diabetic pregnancies was greater than in nondiabefics in each gestational week. Comparison between mean AFI in the diabetic and nondiabetic pregnancies:
Gestational Week
[
15-20 21-25 26-30 31-35 36-41
NonAiabetic 9.3 12.6 12.2 12.0 10.9
[
+ 2.6 -+ 1.8 + 2 -+ 2.4 -+ 3.0
441
Diabetic 13.1 15.0 16.5 16.0 14.8
Table 1. Distribution of AmnioUc Fluid Volume in Twins
Method of Determination
_+ 1.7" • 4.3* _+ 3.5* + 4.0* _+ 4.3*
Dye Dilution Twin l
(*p < 0.00001) CONCLUSION: Gestational age specific values of the amniotic fluid index may be used as a threshold for abnormality. In contrast, no significant difference in AFI values was found between gestational and pregestational diabetes throughout the study period (muhivariant analysis). Moreover, no correlation was found between level of glyeenria (mean 103 4- 19; premeal 93 • 16, postmea1110 -+ 18; Hg-Alc 5.5 -+ 1.4) and AFI at each period of the study.
440
DO MULTIPLE EVALUATIONS IMPROVE THE ACCURACY OF ULTRASOUND AMNIOTIC FLUID VOLUME ASSESSMENT? E.F. Magann, S.P. Chauhan, ~S. Whitwerth,~ A.K. Saltzman,X.]iC Manison. Dept. of O b / Gyn, University of Mississippi, Jackson, MS. OBJECTIVE: To determine if the accuracy of ultrasound estimates of amniotic fluid (AF) volume is improved by using muhiple assessment techniques. STUDY DESIGN: The AF volume of 66 singleton pregnancies was estimated ultrasonically. The ultrasound of each patient was evaluated by subjective assessment (SUBJ, ultrasonic visualization without measurements), the largest vertical pocket (LVP) procedure, the anmiotic fluid index (AFI) and the two-diameter pocket (2DP) technique. The actual AF volume was determined by amniocentesis and a dye-dilution procedure. RESULTS: Based on dye-determined AF volumes and standard criteria, 10 of the pregnancies were classified as oligohydramniotic, 39 normal, and 17 hydramniotic. Table 1 shows the percent of correct ultrasound estimates of AF volume and the predictive value of a normal ultrasound result (PVN).
THE ACCURACY OF THE SUMMATED AMNIOTIC FLUID INDEX T O EVALUATE AMNIOTIC FLUID VOLUME IN UNCOMPLICATED TWIN PREGNANCIES. E./~I Magann, S.P. Chauhan, N.S. lIq~itworth~,~H. Klausen~, B.G. Nevils~,jiC Morrison. Dept. Ob/Gyn, Univ. of Miss., Jackson, MS. OBJECTIVE: To determine the accuracy of the sunmmted amniotic fluid index (sAFI) in the evaluation of the anmiotic fluid (AF) volume in normal twin pregnancies. STUDY DESIGN: The sAFI was determined in 62 normal dianmiotic twin pregnancies by adding ultrasound measures of the deepest vertical pockets in 4 quadrants defined by the umbilicus and linea nigra. Actual AF volume was ascertained in all 124 amniotic sacs by amniocentesis and a dye dilution technique. For data analysis AF volumes were classified as high (> 95% tile), normal (5-95% tile), and low (<5% tile) using previously reported norms. RESULTS: Table 1 compares the distribution olAF volumes as estimated by the sAFI and the actual volume as determined by dye dilution (p < 0.001).
High High Normal Normal Low
Oligo Normal Hydramnios PVN
L VP
AH
2DP
41% (7/17) 80% (31/39) 48% (13/27) 69%
12 (2/17) 92% (36/39) 15% (4/27) 61%
12% (2/17) 95% (35/39) 18% (5/27) 63%
53% (9/17) 85% (33/39) 44% (12/27) 69%
The accuracy of ultrasound estimates of abnormal AF volume were relatively low, ranging fi-orn 12-53% correct. Evaluation of combined ratings of the ultrasound techniques yielded PVNs that varied from a low of 62% (LVP + AFI combined) to a high of 74% (SUBJ + 2DP) combined, p - 0.23). The latter PVN does not ditfer substantially from that of either the SUBJ (69%) or the 2DP (69%) procedures used separately. CONCLUSIONS: Combining the resuhs of different ultrasound techniques does not appear to greatly improve the accuracy of estimates of normal AF volume. Ultrasound estimates of abnormal AF volume have low accuracy. The relatively simple subjective ultrasound estimates of AF volume may be just as accurate as the more complex techniques using ultrasound measurements.
Twin 2
Nermal
Normal Low Normal Low Low
5 4 32 11 10
1 0 0 0 0
4 4 32 10 8
~
Low O 0 O l 2
Estilnates by- the sAFI indicated that AF volumes were normal in 94% (58/62) of the twin pairs. Actual AF volumes, however, were within the normal range for only 52% (32/62) of the sets of twins. The AF volume of 32% (20/62) of the twin pairs showed discordant volumes with one of the pair having an abnormal volume. The sAFI estimated 90% of these discordant sets to be normal. The sensitivity and specificity of the sAFI (normal vs abnormal) was 13 and 100%, respectively. CONCLUSIONS: The sAFI is a poor predictor of low and high AF volumes in uncomplicated twin pregnancies and cannot be relied on to identit~r twin pairs at risk for olighydramnios or hydramnios.
442
AMNIOTIC FLUID INDEX AS A PREDICTOR OF NEONATAL OUTCOME IN PATIENTS WITH PROLONGED PREGNANCY. C O'ReillyGreen, MYDivon. Dept. Of Ob/Gyn, Albert Einstein College of Medicine, Bronx, NY. OBJECTIVE: To evaluate the amniotic fluid index (AFI) as a predictor of neonatal ontcolne. STUDY DESIGN: 625 patients with prolonged pregnancies who had AFI measured 4 days or less before admission were evaluated for neonatal outcome. Variables studied included 1 & 5 minute Apgar scores -< 6 & 7 respectively (1M1N -< 6, 5MIN ~ 7), neonatal intensive care unit admission (NICU), neonatal hospitalization greater than 4 days (>4 DAYS), and any serious neonatal complication (COMPLIC), including resuscitation. Odds ratios were calculated fi-on, ehi-square analysis of 2 • 2 tables, taking an AFI < 5cm as the cutoff for an abnormal test. RESULTS:
True False True False Sens Odds Pos Pos Neg Neg Spec Ratio
Table 1. Correct AF Volume Estimates
SUBJ
J
ummated AFI
1MIN <-- 6 5MIN --< 7 NICU >4DAYS COMPLIC
14 6 7 15 16
109 117 115 39 106
465 495 480 319 459
36 6 20 96 41
.28 .50 .26 .14 .28
.81 .81 .81 .89 .81
1.66 4.23 1.46 1.28 1.69
(95 % C1) (0.80, 3.28) (1.11,16.08) (0.51, 3.70) (0.63, 2.49) (0.85, 3.21)
(POS = positive; NEG = negative, SENS = sensitivity; SPEC = specificity; CI = confidence interval). There were no fi~tal or neonatal deaths in this series, and all neonates were discharged without evidence of impairment. CONCLUSION: 5 minute Apgar score is significantly lower in patients with AFI --< 5, measured 4 days or less before admission. Although odds ratios for other measures of neonatal outcome included 1.0 and thus were not significant, this may represent a Type lI error. Nevertheless, neonatal outcome was excellent for these patients with prolonged pregnancies monitored with AFI as a part of antepartmn surveillance. The relative absence of significant neonatal morbidity is probably related to the delivery of patients with low AFI prior to the development of asphyxia. Prospective, randomized studies are needed in order to establish the role of immediate delivery in the management of oligohydranmios in patients with prolonged pregnancies.