- Email: [email protected]
Amniotic fluid index vs single deepest pocket technique during modified biophysical profile: a randomized clinical trial
American Journal of Obstetrics and Gynecology (2004) 191, 661e8
www.elsevier.com/locate/ajog
Amniotic fluid index vs single deepest pocket technique during modified biophysical profile: A randomized clinical trial Suneet P. Chauhan, MD,a,* Dorota D. Doherty, PhD,b Everett F. Magann, MD,c Francis Cahanding, RN,d Frank Moreno, MD,d Jack H. Klausen, MDd Spartanburg Regional Medical Center, Spartanburg, SCa; Department of Obstetrics and Gynecology, University of Western Australia, Perth, Australiab; Department of Obstetrics and Gynecology, Portsmouth Naval Hospital, Portsmouth, Vac; and Naval Hospital Camp Pendleton, Camp Pendleton, Calif d Received for publication April 13, 2004; revised May 26, 2004; accepted June 18, 2004
KEY WORDS Modified biophysical profile Amniotic fluid Oligohydramnios
Objective: The purpose of this study was to determine the superior technique, if either, of the amniotic fluid index (AFI) vs the single deepest pocket technique in predicting an adverse pregnancy outcome among high-risk patients undergoing antenatal testing. Study design: Patients having modified biophysical profile (nonstress test plus sonographic estimation of amniotic fluid) were randomized to either have AFI or determination of the presence or absence of a 2!1-cm single deepest pocket. Results: Between January of 1997 and December of 2001, 1080 women were randomized with 530 women in the AFI arm, and 558 in the 2!1 pocket arm. The maternal demographics and prenatal complications were similar between groups. Significantly more patients were identified as having oligohydramnios using AFI (17%) compared with using 2!1 pocket (10%) (P = .002). The overall rate of cesarean section for nonreassuring fetal heart rate (FHR) tracing was 3.8% (30 cases, with 16 cases in the AFI-monitored, and 14 cases in the 2!1 pocketemonitored groups, respectively, P = .608). Logistic regression analysis showed no difference between the groups with respect to the ability to identify patients who underwent cesarean section for nonreassuring FHR tracing during labor (P = .999). The umbilical artery pH !7.1 (P = .688) and admission to the newborn intensive care unit were also comparable between groups. Conclusion: During antepartum fetal surveillance, use of single deepest pocket compared with amniotic fluid index is associated with a significantly lower rate of suspected oligohydramnios. Ó 2004 Elsevier Inc. All rights reserved.
Presented at the Sixty-Sixth Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Boca Raton, Florida, January 18-21, 2004. * Reprint requests: Suneet P. Chauhan, MD, Division of MaternalFetal Medicine, Regional Women’s Health Care, 853 N Church St, Suite, Spartanburg, SC 29303. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2004.06.078
Pregnancies at risk for adverse outcomes are managed with modified biophysical profile (BPP) that consists of a nonstress test and sonographic evaluation of amniotic fluid. Antenatal assessment of amniotic fluid is important because it reflects long-term uteroplacental function, and aberrations of this volume have been
662 associated with adverse outcomes.1 Oligohydramnios, for example, is linked with genitourinary anomalies, postural deformities, nonreassuring fetal heart rate (FHR) tracing, and an increased risk of perinatal morbidity and death.2,3 Presently, the 2 techniques to ensure adequacy of amniotic fluid include amniotic fluid index (AFI), which is the summation of the largest vertical pocket in 4 quadrants, and the single deepest pocket technique. According to these 2 estimates of amniotic fluid volume, an AFI %5.0 cm,4 or the absence of a single deepest pocket of at least 2!1 cm is diagnostic of oligohydramnios.5 The AFI is increasingly used in the modified BPP, despite the fact that the ideal cutoff for intervention has not been clearly established. There are conflicting reports about the ability of index % vs O5.0 cm to identify adverse peripartum outcomes,3,6-10 and a paucity of randomized clinical trials. A PubMed search for randomized clinical trials in English with terms ‘‘modified biophysical profile, amniotic fluid index’’ found 3 publications,11-13 and only one of them allocated patients to 1 of the 2 methods of assessing amniotic fluid. Alfirevic et al11 randomized post-term pregnancies (at least 290 days) to AFI and single deepest pocket, and noted the use of AFI was associated with a significantly higher rate of induction without an improvement in perinatal outcomes. Because this publication was limited to postdate pregnancies, and used gestational ageebased criteria for oligohydramnios, we wanted to determine if the findings were applicable to all high-risk pregnancies1 that underwent modified BPP testing. The purpose of this randomized clinical trial was to determine the impact, if any, of using the AFI compared with the single deepest pocket in conjunction with a nonstress test among high-risk pregnancies.
Material and methods This prospective clinical trial was approved by the Clinical Investigation Division at the Naval Medical Center, San Diego, Calif, which is the Institutional Review Board for Naval Hospital Camp Pendleton, and the Investigational Review Board at the University of Mississippi Medical Center. All consecutive high-risk patients undergoing serial fetal surveillance were approached to participate in the prospective, randomized, clinical trial. The inclusion criteria were medical or obstetric problems complicating pregnancy and being managed with weekly modified BPP. The patients were excluded if they had multiple gestations, known fetal anomalies, or refused study participation. If they agreed to take part in the study, an opaque sealed envelope was opened, and the patient was assigned to have the amniotic fluid assessed either with AFI or single deepest pocket.
Chauhan et al As described by Phelan et al, 4 the AFI was obtained by dividing the abdomen in 4 quadrants, obtaining the deepest pocket in each quadrant and summing them up. Patients assigned to the single deepest pocket were evaluated according to Chamberlain et al.14 Oligohydramnios defined as AFI %5.0 cm, or absence of a single deepest pocket of 2!1 cm was an indication for delivery if the gestational age was 34 weeks or more. Once a patient was assigned to AFI or single deepest pocket arm, all subsequent assessments of amniotic fluid were done accordingly. Clinicians managing the patients were aware of the results of the modified BPP, and whether AFI or single deepest pocket was used to evaluate the amniotic fluid. The amniotic fluid estimation, AFI or single deepest pocket, used for this evaluation is the last volume estimated before delivery. In pregnancies complicated by diabetes or intrauterine growth restriction the last amniotic fluid estimation would have been within 3 days of delivery, and in all other pregnancies within 7 days of delivery. All cesarean deliveries were done by obstetrics and gynecology residents under the supervision of an attending. Cesarean delivery for cephalopelvic disproportion was done according to the American College of Gynecology guidelines,15 and they were done for nonreassuring FHR tracing if the abnormality did not respond to intrauterine resuscitative efforts like hydration, position change, oxygenation, amnioinfusion, and terbutaline, where applicable.16 At delivery, the umbilical cord was doubly clamped, and an arterial blood sample was obtained in a preheparinized syringe. The sample was placed on ice and analyzed (Corning 178 analyzer, Medfield, Mass) within 30 minutes of the delivery. Randomization was accomplished via use of a computer-generated random number table with blocked permutations. Randomization was accomplished by opening sealed opaque envelopes containing group allocations that were prepared in blocks of 14 envelopes, 7 per allocation. When the envelope pack was reduced to 8 envelopes, a new block of 14 envelopes was supplemented. A person not directly associated with the study performed randomization and envelope preparation. The primary end point of the analysis was the incidence of cesarean section for nonreassuring FHR tracing. Sample size calculations indicated that a sample of 1080 women would achieve 80% power in detecting a difference of 5% vs 9% in the rate of cesarean section for nonreassuring FHR tracing between treatment arms using a chi-square test at 5% significance level. Descriptive statistics used means and standard deviations or medians and interquartile ranges (Q1-Q3) and percentages, as appropriate for continuous and categorical data. Performance of AFI and 2!1 pocket monitoring was also evaluated by their sensitivity (proportion of cases with adverse outcome identified with either AFI
Chauhan et al or 2!1 pocket monitoring) and specificity (proportion of normal outcomes as identified with normal amniotic fluid volume using either AFI or 2!1 pocket monitoring) to identify women likely to experience adverse outcomes. The primary end point and other categorical outcomes were investigated using chi-square tests. All subjects were analyzed on an ‘‘intent to treat’’ basis. Supplementary analyses included logistics regression modeling to investigate factors associated with each outcome considered (incidence of nonreassuring FHR tracing, cesarean section for nonreassuring FHR tracing, and admission to neonatal intensive care nursery). In logistic regression models, an interaction of mode of management (AFI vs 2!1 pocket) and ‘high-risk’ status (defined as AFI %5.0 cm or 2!1 pocket not present) was investigated in combination to assess whether one of these indices was more predictive of adverse outcome. Along with sensitivity and specificity, likelihood ratios (LR = sensitivity over 1- specificity) for positive (oligohydramnios) and negative (adequate amniotic fluid) tests were calculated for each group. According to the guidelines established by the Evidence-Based Medicine Working Group,17 a diagnostic test may be categorized as definitively useful if LR O10 or !0.1, moderately useful if LR 5 to 10 or 0.1 to 0.2, slightly useful if LR 2 to 5 or 0.2 to 0.5, and not useful LR 1 to 2 or 0.5 to 1. Statistical analysis was conducted using SPSS statistical software (SPSS version 11.0, Inc, Chicago, Ill). Logistic regression analyses were carried out using LogXact logistic regression software (Cytel version 5, Cambridge, Mass). P values ! .05 were considered statistically significant.
Results One thousand eighty women were randomized into 1 of the 2 management regimens, using either AFI or 2!1 pocket assessment antenatally. Five hundred fifty-eight women were managed with the 2!1 pocket, and 530 patients were monitored with the AFI. Maternal demographics were similar between the groups (Table I). Peripartum outcomes are summarized in Table II. There was a considerable difference between the ‘high-risk’ status as defined by AFI %5.0 cm compared with the ‘high-risk’ status assigned due to the absence of 2!1 pocket (88 patients [17%] vs 57 patients [10%]). Within the preterm deliveries (gestational age !37 weeks), 19% (9 cases) had oligohydramnios in the AFI monitored group and 11% (6 cases) in the 2!1 pocket group (P = .401). There were no differences between the rates of variable decelerations present (P = .662); moreover, there were no differences in variable decelerations influencing delivery among the patients where variable decelerations
663 Table I Maternal characteristics stratified by mode of monitoring, AFI and 2!1 pocket Group Characteristic
AFI (n = 530) 2!1 (n = 558) P value
Maternal age 24 Race Caucasian 317 African American 104 Others 109 Primigravida 185 Nulliparous 265 Prenatal complications Hypertension 64 Diabetes 68 IUGR 39 Raised MSAFP 66 Post dates 204 Oligohydramnios 13 Other 76 40 Gestational age at delivery Gestational age distribution !37 47 37-40 264 Op40 219
(IQ 21-30) 25 (IQ 21-29) 0.396 (60%) (20%) (20%) (35%) (50%)
337 104 116 202 281
(12%) 63 (13%) 67 (7%) 42 (13%) 92 (39%) 210 (3%) 8 (14%) 76 (IQ 38-41) 40
(9%) (50%) (41%)
(60%) (19%) (21%) (36%) (50%)
0.924 0.656 0.906
(11%) (12%) (8%) (17%) (38%) (1%) (14%) 0.548 (IQ 38-41) 0.561
53 (10%) 279 (50%) 226 (41%)
0.921
were observed (P = .279). There were differences between the rates of minimal or absent variability present between the groups (P = .046). Late decelerations were similar between the groups (P = .133), and the rates of abnormal decelerations influencing delivery were also alike (P = .279). Mode of delivery and reasons for operative delivery were alike (Table II). The rate of nonreassuring FHR tracing was similar between the groups (P = .390), and logistic regression analysis showed no difference between the groups with respect to the ability of the sonographic estimation of amniotic fluid to identify patients with nonreassuring FHR tracing during labor (P = .731). The overall rate of cesarean section for nonreassuring FHR tracing was 3.8% (30 cases with 16 cases in the AFI-monitored, and 14 in the 2!1 pocketemonitored groups, P = .608), and these cesarean sections comprised 43% of all sections in the AFI-monitored group (out of 37 cesarean deliveries) and 38% of all cesarean sections in the 2!1-monitored group (out of 37 c/s) (P = .637). Logistic regression analysis showed no difference between the groups with respect to the ability to identify patients who underwent cesarean section for nonreassuring FHR tracing during labor (P = .999). There were no differences between the groups for most of the neonatal outcomes (Table III). There was no difference in the rate of admission to the newborn ICU
664 Table II
Chauhan et al Peripartum outcomes stratified by sonographic assessment of AF volume, using either AFI or 2!1 pocket Group
Characteristic Oligohydramnios* Amnioinfusion Nonreassuring fetal heart rate (FHR) tracingy Nonreactive NST Variable decelerations Present Infl deliveryy Variability Min/absent Late decelerations Present Infl deliveryy Non-reassuring fetal heart rate (FHR) tracing Delivery mode Spont. Vaginal Assist. Vaginal Cesarean Reason for cesarean deliveryy Nonreassuring fetal heart rate (FHR) tracing Failure to progress Other Reason for assisted vaginal deliveryy,z Nonreassuring fetal heart rate (FHR) tracing Failure to progress Other
AFI (n = 530) 88 56 34 44
(17%) (11%) (63%) (8%)
2!1 (n = 558) 57 44 21 51
P value
(10%) (8%) (48%) (9%)
0.002 0.126 0.155 0.668
77 (15%) 67 (87%)
75 (13%) 60 (80%)
0.662 0.279
41 (8%)
63 (11%)
0.046
37 (7%) 33 (94%) 87 (16%)
27 (5%) 26 (96%) 83 (15%)
0.133 0.999 0.390
365 (69%) 128 (24%) 37 (7%)
405 (73%) 116 (21%) 37 (7%)
0.377
16 (43%) 7 (19%) 13 (35%)
14 (38%) 14 (38%) 9 (24%)
0.766
71 (56%) 44 (34%) 13 (10%)
69 (60%) 35 (30%) 11 (10%)
0.213
* AFI %5.0 cm in the AFI-monitored group, or 2!1 pocket not present in the 2!1 group, on the intrapartum AF assessment. y Percentages for subsets of patients with condition present are shown. One case of missing information about reason for cesarean delivery in AFI-monitored group, and one case of missing reason for assisted vaginal delivery in the 2!1 pocketemonitored group. z Three cases of nonreassuring fetal heart rate (FHR) tracing noted in spontaneous deliveries (2 in AFI-monitored group and 1 in the 2!1 pocket group).
(P = .328), but among the neonates admitted to the newborn ICU there was a higher proportion of admission for respiratory distress in the AFI-monitored group (P = .028). Among the preterm delivery, the rate of respiratory distress syndrome (RDS) was significantly higher among those randomized to the AFI (0.1%; 6/530) than to 2!1 pocket (0%; P = .009). These 6 premature deliveries were results of hypertensive disease (n = 2), preterm labor refractory to treatment with tocolytics (n = 3), and one case of fetal growth restriction and oligohydramnios (AFI of 4.0 cm) at 34 weeks. This last case was induced for abnormal growth and inadequate amniotic fluid, and had an uncomplicated intrapartum course with a vaginal delivery of a newborn weighing 1950 g, Apgar score of 5 at 5 minutes, and umbilical arterial pH of 7.20. Thus, the rate of iatrogenic prematurity due to the use of AFI was 2% (1/47; 95% CI 0.05e11). There were no cases of stillbirths in either group. The predictive accuracies of the 2 methods of assessing amniotic fluid are provided in Table IV. Both AFI and 2!1 monitoring methods were characterized by low
sensitivity to identify women likely to experience adverse outcomes with relatively high ability to predict normal outcomes. Use of likelihood ratios and guidelines provided by the Evidence-Based Medicine Working Group17 indicates that both methods of evaluating amniotic fluid are poor diagnostic tests in identifying patients that will undergo cesarean delivery for nonreassuring FHR tracing or deliver a newborn with depression or acidosis.
Comment The results of this randomized clinical trial are notable for 3 major findings. First, during modified BPP, use of the AFI rather than single deepest pocket to estimate the amniotic fluid volume increases the likelihood of intervention without a demonstrable benefit. High-risk patients were significantly more likely to be diagnosed as having oligohydramnios, be induced for an abnormal fluid volume, and undergo amnioinfusion if AFI is used for fetal assessment than the measurement of the largest
Chauhan et al Table III
665
Neonatal characteristics stratified by mode of monitoring, AFI and 2!1 pocket Group
Characteristic
AFI (n = 530)
2!1 (n = 558)
P value
Birth weight Birth weight distribution !1500 1500-2500 2500-4000 O4000 Meconium Present Thicky Umb artery pH !7.1* Apgar score !7 1 min 5 min Admission to NICU Admission reasonsy RDS/TTN Sepsis Other
3490 (3074-3856)
3250 (3105-3830)
0.780
1 25 414 90
(0.2%) (5%) (78%) (17%)
1 25 448 84
(0.2%) (5%) (80%) (15%)
0.843
65 39 14 68 12
(12%) (60%) (3%) (13%) (2%)
64 45 17 67 6
(12%) (70%) (3%) (12%) (1%)
0.685 0. 269 0.688 0.681 0.124
35 27 4 2
(7%) (82%) (12%) (6%)
29 15 11 1
(5%) (56%) (41%) (4%)
0.328 0.419 0.028
* There were 2 cases (ie, 0.4%) of pH !7.00 in each group. y Percentages for subsets of patients with condition present are shown.
Table IV
The predictive accuracies of adverse outcomes using amniotic fluid index or the presence/absence of a 2!1 pocket AFI (n = 530)
CD for distress AS !7 at 5 min UApH !7.10 UApH !7.00
2!1 (n = 558)
Sens
Spec
PPV
NPV
LR (C)
LR ()
Sens
Spec
PPV
NPV
LR (C)
LR ()
6.3% 8.3% 7.1% 50%
83.1% 83.2% 83.1% 83.5%
0.8% 0.6% 0.9% 0.3%
96.6% 97.5% 96.9% 99.8%
0.4 0.5 0.4 3.0
0.9 0.9 0.8 1.6
7.7% 0 5.9% 0
89.7% 89.7% 89.6% 89.7%
1.8% 0 1.8% 0
97.6% 98.8% 96.8% 99.6%
0.7 0 0.6 0
0.9 0.8 0.9 0.8
Sens, Sensitivity; Spec, specificity; PPV, positive predictive value; NPV, negative predictive value .
pocket. Despite the increased rate of intervention, 2 major outcomes of concern, cesarean delivery for nonreassuring FHR and neonatal acidosis, were similar in the 2 groups. Second, because of a false-positive test, iatrogenic prematurity does occur when modified biophysical is used for antepartum surveillance. While we noted its rate to be 2%, Miller et al17 reported that it was 1.5%. Though this complication was similar for the 2 groups, it may be due to the small sample size of this study. We did note a significantly higher rate of RDS among those who had assessment of amniotic fluid using the index rather than the single deepest pocket. The preterm deliveries in 5 of the 6 cases of RDS were secondary to medical (hypertension) or obstetric (spontaneous preterm labor) complications and use of AFI spuriously influenced the rate of RDS. Only one case of preterm delivery and RDS was secondary to the finding of AFI !5.0 cm. Thus, iatrogenic prematurity with associated complications, we think, occurs in 2% (1/47; 95% CI 0.05e11) of the cohorts that are managed with the use of
AFI. Randomized clinical trials with a larger sample size are needed to determine a more precise risk of iatrogenic RDS with use of AFI in the modified BPP. The third finding of this trial is that both methods of determining amniotic fluid are poor diagnostic tests in identifying peripartum complications (Table IV). Presently, there is a debate about the diagnostic ability of AFI to predict adverse outcomes. While Casey et al3 noted that with the use of AFI, antepartum detection of oligohydramnios is associated with increased perinatal morbidity and mortality, Magann et al6 considered AFI to be a poor diagnostic test. The limitation of these reports3,6 is that they were not randomized clinical trials, and the investigators did not report the likelihood ratio, a preferred way to evaluate a diagnostic test. The findings of this study are consistent with the previous report by Alfirevic et al,11 who randomized uncomplicated, post-term patients (gestational age of 290 days or more) to be monitored with computerized cardiotocography with AFI or maximum deepest pocket. These investigators considered the amniotic fluid to be
666 abnormal if the AFI was less than 7.3 cm or if the maximum deepest pocket was !1.8 cm, both of which are below the third percentile for 42 weeks. Despite the differences in the definition of inadequate amniotic fluid between our study and this published report,11 Alfirevic et al’s study also noted that significantly more post-term patients were diagnosed with oligohydramnios, and were subsequently induced if the AFI was used for antepartum surveillance. Both our randomized study and that of Alfirevic et al11 noted that the rate of cesarean delivery for nonreassuring FHR tracing and neonatal acidosis were similar, regardless of which technique was used to estimate the amniotic fluid volume. Amniotic fluid volume assessment is important in the surveillance of high-risk pregnancies. A low amniotic fluid volume is thought to represent inadequate uteroplacental perfusion with the shunting of blood to the fetal heart, adrenals, and brain with resultant poor renal perfusion and a reduced fetal urinary output. Before the investigations by Phelan et al4 suggesting that the AFI be used to evaluate amniotic fluid volume, the single deepest pocket was routinely used, and the definition of oligohydramnios was the absence of a 1!1 or 2!1 cm pocket.5 Surprisingly, without the benefit of a randomized clinical trial, the AFI was substituted for the single deepest pocket technique to estimate amniotic fluid volume. A possible reason for this substitution were the reports by Rutherford et al19 and Sarno et al,20 which noted that AFI %5.0 cm was significantly associated with an increased rate of cesarean delivery for nonreassuring FHR tracing and low Apgar scores. What was overlooked in this substitution was the subjective nature of these 2 end points. The incidence of neonatal acidosis, an objective assessment of fetal well-being, with AFI % vs O5.0 cm was not even addressed in the early reports.19,20 Subsequently, a meta-analysis of 42 reports over 10 years found only one study which correlated the rate of umbilical arterial pH !7.00 with AFI, and noted that neonatal acidosis occurs with similar frequency among high-risk patients with AFI %5.0 cm as those with O5.0 cm.10 More recent works by Magann et al6 and Casey et al3 have also not linked pH !7.00 with AFI O5.0 cm. If there is no association between pathologic acidosis and AFI, then continued use of this technique to assess amniotic fluid is unnecessary, especially because randomized clinical trials indicate an increase rate of intervention without improved outcome. The major shortcoming of this randomized trial is that the sample size was inadequate to evaluate meaningful, unambiguous outcomes like cerebral palsy or neonatal death. But considering the infrequent occurrence of these events, a randomized clinical trial to decrease their rate is a formidable undertaking. Assuming that umbilical arterial pH !7.00 occurs in 0.4%, then 8700 patients in each group are necessary to detect
Chauhan et al a 20% reduction in prevalence between the study arms (two-tailed alpha of 0.05 and power of 80%). It is noteworthy that Manning et al21 did report that with the use of the complete BPP they were able to significantly reduce the rate of cerebral palsy among high-risk patients. Specifically, they noted that the incidence of cerebral palsy among the general population (3.7 per 1000 live births) was significantly higher than the high-risk parturients who had BPP (1.3 per 1000 live births; P ! .0001). Admittedly, several criteria have been used to ensure adequate fluid during the BPPs,5 but none of them included use of AFI. Thus, if modified BPP is to be substituted for the 5 variables survey, then we should use one of the definitions used by Manning et al.21 In conclusion, use of single deepest pocket with modified BPP would decrease the rate of induction for suspected oligohydramnios. Unless randomized clinical trials show benefit of summing the measurements of 4 quadrants over dimensions of the deepest pocket, use of AFI during modified BPP should be avoided.
References 1. American College of Obstetricians and Gynecologists. Antenatal fetal surveillance. Washington (DC): American College of Obstetricians and Gynecologists; 1993. ACOG Practice Bulletin No. 9. 2. Golan A, Lin G, Evron S, Arieli S, Niv D, David MP. Oligohydramnios: maternal complications and fetal outcome in 145 cases. Gynecol Obstet Invest 1994;37:91-5. 3. Casey BM, McIntire DD, Bloom SL, Lucas MJ, Santos R, Twickler DM, et al. Pregnancy outcomes after antepartum diagnosis of oligohydramnios at or beyond 34 weeks’ gestation. Am J Obstet Gynecol 2000;182:909-12. 4. Phelan JP, Smith CV, Broussard P, Small M. Amniotic fluid volume assessment with the four-quadrant technique at 36-42 weeks’ gestation. J Repro Med 1987;32:540-2. 5. Magann EF, Isler CM, Chauhan SP, Martin JN Jr. Amniotic fluid volume estimation and the biophysical profile: a confusion of criteria. Obstet Gynecol 2000;96:640-2. 6. Magann EF, Chauhan SP, Kinsella MJ, McNamara MF, Whitworth NS, Morrison JC. Antenatal testing among 1001 patients at high risk: the role of ultrasonographic estimate of amniotic fluid volume. Am J Obstet Gynecol 1999;180:1330-6. 7. Rainford M, Adair R, Scialli A, Ghidini A, Spong CY. Amniotic fluid index in uncomplicated term pregnancy: prediction of outcome. J Reprod Med 2001;46:589-92. 8. Rutherford SE, Phelan JP, Smith CV, Jacobs N. The fourquadrant assessment of amniotic fluid volume: an adjunct to antepartum fetal heart rate testing. Obstet Gynecol 1987;70:353-6. 9. Voxman EG, Tran S, Wing DA. Low amniotic index as a predictor of adverse perinatal outcome. J Perinatol 2002;22:282-5. 10. Chauhan SP, Sanderson M, Hendrix NW, Magann EF, Devoe LD. Perinatal outcome and amniotic fluid index in the antepartum and intrapartum periods: a meta-analysis. Am J Obstet Gynecol 1999;181:1473-8. 11. Alfirevic Z, Luckas M, Walkinshaw SA, McFarlane M, Curran R. A randomized comparison between amniotic fluid index and maximum pool depth in the monitoring of post-term pregnancy. BJOG 1997;104:207-11. 12. Amano K, Saito K, Shoda T, Tani A, Youshihara H, Nishijima M. Elective induction of labor at 39 weeks of gestation: a prospective randomized trial. J Obstet Gynaecol Res 1999;25:33-7.
Chauhan et al 13. Nageotte MP, Towers CV, Asrat T, Freeman RK. Perinatal outcome with modified biophysical profile. Am J Obstet Gynecol 1994;170:1672-6. 14. 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 volume to perinatal outcome. Am J Obstet Gynecol 1984;150:245-9. 15. American College of Obstetricians and Gynecologists. Dystocia and the augmentation of labor. ACOG technical bulletin 218. Washington, DC: A; 1995. 16. American College of Obstetricians and Gynecologists. Fetal heart rate patterns: Monitoring, interpretation, and management. Washington, DC: American College of Obstetricians and Gynecologists; 1995. ACOG technical bulletin 207. 17. Jaeschke R, Guyatt GH, Sackett DL, for the Evidence-Based Medicine Working Group. Users’ guide to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients?. JAMA 1994;271:703-7. 18. Miller DA, Rabello YA, Paul RH. The modified biophysical profile: antepartum testing in the 1990s. Am J Obstet Gynecol 1996;174:82-7. 19. Rutherford SE, Phelan JP, Smith CV. The four-quadrant assessment of amniotic fluid volume: an adjunct to antepartum fetal heart rate testing. Obstet Gynecol 1987;70:353-6. 20. Sarno AP, Ahn MO, Brar HS, Phelan JP, Platt LD. Intrapartum Doppler velocimetry, amniotic fluid volume, and fetal heart rate as predictors of subsequent non-reassuring fetal heart rate (FHR) tracing: an initial report. Am J Obstet Gynecol 1989;161:1508-14. 21. Manning FA, Bondaji N, Harman CR, Casiro O, Menticoglou S, Morrison I, et al. Fetal assessment based on fetal biophysical profile scoring.VIII. Am J Obstet Gynecol 1998;178:696-706.
Discussion DR WILLIAM A. NEWTON, Tampa, Fla. Oligohydramnios has been demonstrated by a number of studies as an ominous prognostic sign of fetal well-being, especially in the high-risk and postdate pregnancy.1-3 There have been links between decreased amniotic fluid volume to stillbirth, fetal anomaly, abnormal fetal heart rate tracings in labor, increase in cesarean section for fetal distress, and possibly fetal acidosis.4 The concept of 4quadrant fluid measurement for amniotic fluid volume (AFV) was introduced in the mid 1980s. Since that time amniotic fluid index (AFI) has been widely accepted as an accurate means of defining oligohydramnios and its relationship to fetal compromise.2 However, in the past several years AFI has increasingly been questioned as the appropriate standard measurement to define oligohydramnios.5 Because many authors have noted that amniotic fluid volume alone may be a poor predictor of fetal wellbeing6 there has been the evolution of amniotic fluid volume incorporated into other diagnostic tests such as the biophysical profile and the modified biophysical profile. These fetal assessment tests greatly enhance safe management of the high-risk pregnancy but may also result in unwarranted intervention and morbidity.
667 Dr Chauhan’s study compares AFI versus measurement of a 2!1-cm single deepest pocket (SDP) incorporated into the modified biophysical profile as a predictor of fetal compromise. His results imply that AFI is not only less specific than SDP but more likely to result in a higher potential for unnecessary intervention through increased induction of labor for oligohydramnios. Using AFI with modified biophysical profile resulted in a 2% iatrogenic prematurity rate in this study. These findings seem contradictory to the 1987 study by Rutherford et al,7 which looked at 330 highrisk pregnancies comparing 4-quadrant AFV assessment and SDP. In that study AFI was found to have a high relationship to low Apgar score and cesarean section for fetal distress. Following Rutherford’s study the AFI was used as measurement of amniotic fluid volume for the modified biophysical profile. Coinciding with the acceptance of AFI as routine part of obstetric ultrasound, induction of labor for oligohydramnios increased from 26% in 1991 to 75% in 1996. In Dr Chauhan’s study, intrapartum oligohydramnios was diagnosed in 17% by AFI and only 10% by SDP. However, there was no significant difference in predictability of cesarean section for fetal distress, low Apgar score, or umbilical pH !7.1. Of note, clinicians managing labor were aware of whether oligohydramnios was determined by AFI or SDP. There was also the implication that AFI measurement was associated with increased potential morbidity, as noted by 13 more AFI patients undergoing amnioinfusion for presumed fetal distress. Admission to the NICU for transient tachypnea of the newborn and respiratory distress syndrome occurred in 12 more patients with AFI vs SDP, implying that in a larger sample size, AFI may be statistically associated with a higher iatrogenic morbidity than SDP. AFI as well as SDP are ‘‘poor diagnostic tests’’ for identifying peripartum complications. If used in the modified biophysical profile, the SDP should decrease the rate of induction of labor for suspected oligohydramnios. This study is impressive, but a much larger sample size is needed to look at the best method of determining oligohydramnios. For that matter, defining exactly what oligohydramnios is quantitatively and in what circumstance it reflects a poor prognostic sign is needed. The modified biophysical profile with assessment of AFV is an easy and convenient test for the evaluation of the high-risk patient. A reliable consensus standard of measurement is essential for a uniform diagnostic test. Dr Chauhan et al are congratulated on this interesting prospective clinical study. My questions for Dr Chauhan are: (1) in your study, the managing clinicians were aware of the method of determining oligohydramnios. Don’t you feel that this may have resulted in clinician bias in the management of individual cases? (2) What role, if any, do you feel that evolving technology
www.elsevier.com/locate/ajog
Amniotic fluid index vs single deepest pocket technique during modified biophysical profile: A randomized clinical trial Suneet P. Chauhan, MD,a,* Dorota D. Doherty, PhD,b Everett F. Magann, MD,c Francis Cahanding, RN,d Frank Moreno, MD,d Jack H. Klausen, MDd Spartanburg Regional Medical Center, Spartanburg, SCa; Department of Obstetrics and Gynecology, University of Western Australia, Perth, Australiab; Department of Obstetrics and Gynecology, Portsmouth Naval Hospital, Portsmouth, Vac; and Naval Hospital Camp Pendleton, Camp Pendleton, Calif d Received for publication April 13, 2004; revised May 26, 2004; accepted June 18, 2004
KEY WORDS Modified biophysical profile Amniotic fluid Oligohydramnios
Objective: The purpose of this study was to determine the superior technique, if either, of the amniotic fluid index (AFI) vs the single deepest pocket technique in predicting an adverse pregnancy outcome among high-risk patients undergoing antenatal testing. Study design: Patients having modified biophysical profile (nonstress test plus sonographic estimation of amniotic fluid) were randomized to either have AFI or determination of the presence or absence of a 2!1-cm single deepest pocket. Results: Between January of 1997 and December of 2001, 1080 women were randomized with 530 women in the AFI arm, and 558 in the 2!1 pocket arm. The maternal demographics and prenatal complications were similar between groups. Significantly more patients were identified as having oligohydramnios using AFI (17%) compared with using 2!1 pocket (10%) (P = .002). The overall rate of cesarean section for nonreassuring fetal heart rate (FHR) tracing was 3.8% (30 cases, with 16 cases in the AFI-monitored, and 14 cases in the 2!1 pocketemonitored groups, respectively, P = .608). Logistic regression analysis showed no difference between the groups with respect to the ability to identify patients who underwent cesarean section for nonreassuring FHR tracing during labor (P = .999). The umbilical artery pH !7.1 (P = .688) and admission to the newborn intensive care unit were also comparable between groups. Conclusion: During antepartum fetal surveillance, use of single deepest pocket compared with amniotic fluid index is associated with a significantly lower rate of suspected oligohydramnios. Ó 2004 Elsevier Inc. All rights reserved.
Presented at the Sixty-Sixth Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Boca Raton, Florida, January 18-21, 2004. * Reprint requests: Suneet P. Chauhan, MD, Division of MaternalFetal Medicine, Regional Women’s Health Care, 853 N Church St, Suite, Spartanburg, SC 29303. E-mail: [email protected] 0002-9378/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2004.06.078
Pregnancies at risk for adverse outcomes are managed with modified biophysical profile (BPP) that consists of a nonstress test and sonographic evaluation of amniotic fluid. Antenatal assessment of amniotic fluid is important because it reflects long-term uteroplacental function, and aberrations of this volume have been
662 associated with adverse outcomes.1 Oligohydramnios, for example, is linked with genitourinary anomalies, postural deformities, nonreassuring fetal heart rate (FHR) tracing, and an increased risk of perinatal morbidity and death.2,3 Presently, the 2 techniques to ensure adequacy of amniotic fluid include amniotic fluid index (AFI), which is the summation of the largest vertical pocket in 4 quadrants, and the single deepest pocket technique. According to these 2 estimates of amniotic fluid volume, an AFI %5.0 cm,4 or the absence of a single deepest pocket of at least 2!1 cm is diagnostic of oligohydramnios.5 The AFI is increasingly used in the modified BPP, despite the fact that the ideal cutoff for intervention has not been clearly established. There are conflicting reports about the ability of index % vs O5.0 cm to identify adverse peripartum outcomes,3,6-10 and a paucity of randomized clinical trials. A PubMed search for randomized clinical trials in English with terms ‘‘modified biophysical profile, amniotic fluid index’’ found 3 publications,11-13 and only one of them allocated patients to 1 of the 2 methods of assessing amniotic fluid. Alfirevic et al11 randomized post-term pregnancies (at least 290 days) to AFI and single deepest pocket, and noted the use of AFI was associated with a significantly higher rate of induction without an improvement in perinatal outcomes. Because this publication was limited to postdate pregnancies, and used gestational ageebased criteria for oligohydramnios, we wanted to determine if the findings were applicable to all high-risk pregnancies1 that underwent modified BPP testing. The purpose of this randomized clinical trial was to determine the impact, if any, of using the AFI compared with the single deepest pocket in conjunction with a nonstress test among high-risk pregnancies.
Material and methods This prospective clinical trial was approved by the Clinical Investigation Division at the Naval Medical Center, San Diego, Calif, which is the Institutional Review Board for Naval Hospital Camp Pendleton, and the Investigational Review Board at the University of Mississippi Medical Center. All consecutive high-risk patients undergoing serial fetal surveillance were approached to participate in the prospective, randomized, clinical trial. The inclusion criteria were medical or obstetric problems complicating pregnancy and being managed with weekly modified BPP. The patients were excluded if they had multiple gestations, known fetal anomalies, or refused study participation. If they agreed to take part in the study, an opaque sealed envelope was opened, and the patient was assigned to have the amniotic fluid assessed either with AFI or single deepest pocket.
Chauhan et al As described by Phelan et al, 4 the AFI was obtained by dividing the abdomen in 4 quadrants, obtaining the deepest pocket in each quadrant and summing them up. Patients assigned to the single deepest pocket were evaluated according to Chamberlain et al.14 Oligohydramnios defined as AFI %5.0 cm, or absence of a single deepest pocket of 2!1 cm was an indication for delivery if the gestational age was 34 weeks or more. Once a patient was assigned to AFI or single deepest pocket arm, all subsequent assessments of amniotic fluid were done accordingly. Clinicians managing the patients were aware of the results of the modified BPP, and whether AFI or single deepest pocket was used to evaluate the amniotic fluid. The amniotic fluid estimation, AFI or single deepest pocket, used for this evaluation is the last volume estimated before delivery. In pregnancies complicated by diabetes or intrauterine growth restriction the last amniotic fluid estimation would have been within 3 days of delivery, and in all other pregnancies within 7 days of delivery. All cesarean deliveries were done by obstetrics and gynecology residents under the supervision of an attending. Cesarean delivery for cephalopelvic disproportion was done according to the American College of Gynecology guidelines,15 and they were done for nonreassuring FHR tracing if the abnormality did not respond to intrauterine resuscitative efforts like hydration, position change, oxygenation, amnioinfusion, and terbutaline, where applicable.16 At delivery, the umbilical cord was doubly clamped, and an arterial blood sample was obtained in a preheparinized syringe. The sample was placed on ice and analyzed (Corning 178 analyzer, Medfield, Mass) within 30 minutes of the delivery. Randomization was accomplished via use of a computer-generated random number table with blocked permutations. Randomization was accomplished by opening sealed opaque envelopes containing group allocations that were prepared in blocks of 14 envelopes, 7 per allocation. When the envelope pack was reduced to 8 envelopes, a new block of 14 envelopes was supplemented. A person not directly associated with the study performed randomization and envelope preparation. The primary end point of the analysis was the incidence of cesarean section for nonreassuring FHR tracing. Sample size calculations indicated that a sample of 1080 women would achieve 80% power in detecting a difference of 5% vs 9% in the rate of cesarean section for nonreassuring FHR tracing between treatment arms using a chi-square test at 5% significance level. Descriptive statistics used means and standard deviations or medians and interquartile ranges (Q1-Q3) and percentages, as appropriate for continuous and categorical data. Performance of AFI and 2!1 pocket monitoring was also evaluated by their sensitivity (proportion of cases with adverse outcome identified with either AFI
Chauhan et al or 2!1 pocket monitoring) and specificity (proportion of normal outcomes as identified with normal amniotic fluid volume using either AFI or 2!1 pocket monitoring) to identify women likely to experience adverse outcomes. The primary end point and other categorical outcomes were investigated using chi-square tests. All subjects were analyzed on an ‘‘intent to treat’’ basis. Supplementary analyses included logistics regression modeling to investigate factors associated with each outcome considered (incidence of nonreassuring FHR tracing, cesarean section for nonreassuring FHR tracing, and admission to neonatal intensive care nursery). In logistic regression models, an interaction of mode of management (AFI vs 2!1 pocket) and ‘high-risk’ status (defined as AFI %5.0 cm or 2!1 pocket not present) was investigated in combination to assess whether one of these indices was more predictive of adverse outcome. Along with sensitivity and specificity, likelihood ratios (LR = sensitivity over 1- specificity) for positive (oligohydramnios) and negative (adequate amniotic fluid) tests were calculated for each group. According to the guidelines established by the Evidence-Based Medicine Working Group,17 a diagnostic test may be categorized as definitively useful if LR O10 or !0.1, moderately useful if LR 5 to 10 or 0.1 to 0.2, slightly useful if LR 2 to 5 or 0.2 to 0.5, and not useful LR 1 to 2 or 0.5 to 1. Statistical analysis was conducted using SPSS statistical software (SPSS version 11.0, Inc, Chicago, Ill). Logistic regression analyses were carried out using LogXact logistic regression software (Cytel version 5, Cambridge, Mass). P values ! .05 were considered statistically significant.
Results One thousand eighty women were randomized into 1 of the 2 management regimens, using either AFI or 2!1 pocket assessment antenatally. Five hundred fifty-eight women were managed with the 2!1 pocket, and 530 patients were monitored with the AFI. Maternal demographics were similar between the groups (Table I). Peripartum outcomes are summarized in Table II. There was a considerable difference between the ‘high-risk’ status as defined by AFI %5.0 cm compared with the ‘high-risk’ status assigned due to the absence of 2!1 pocket (88 patients [17%] vs 57 patients [10%]). Within the preterm deliveries (gestational age !37 weeks), 19% (9 cases) had oligohydramnios in the AFI monitored group and 11% (6 cases) in the 2!1 pocket group (P = .401). There were no differences between the rates of variable decelerations present (P = .662); moreover, there were no differences in variable decelerations influencing delivery among the patients where variable decelerations
663 Table I Maternal characteristics stratified by mode of monitoring, AFI and 2!1 pocket Group Characteristic
AFI (n = 530) 2!1 (n = 558) P value
Maternal age 24 Race Caucasian 317 African American 104 Others 109 Primigravida 185 Nulliparous 265 Prenatal complications Hypertension 64 Diabetes 68 IUGR 39 Raised MSAFP 66 Post dates 204 Oligohydramnios 13 Other 76 40 Gestational age at delivery Gestational age distribution !37 47 37-40 264 Op40 219
(IQ 21-30) 25 (IQ 21-29) 0.396 (60%) (20%) (20%) (35%) (50%)
337 104 116 202 281
(12%) 63 (13%) 67 (7%) 42 (13%) 92 (39%) 210 (3%) 8 (14%) 76 (IQ 38-41) 40
(9%) (50%) (41%)
(60%) (19%) (21%) (36%) (50%)
0.924 0.656 0.906
(11%) (12%) (8%) (17%) (38%) (1%) (14%) 0.548 (IQ 38-41) 0.561
53 (10%) 279 (50%) 226 (41%)
0.921
were observed (P = .279). There were differences between the rates of minimal or absent variability present between the groups (P = .046). Late decelerations were similar between the groups (P = .133), and the rates of abnormal decelerations influencing delivery were also alike (P = .279). Mode of delivery and reasons for operative delivery were alike (Table II). The rate of nonreassuring FHR tracing was similar between the groups (P = .390), and logistic regression analysis showed no difference between the groups with respect to the ability of the sonographic estimation of amniotic fluid to identify patients with nonreassuring FHR tracing during labor (P = .731). The overall rate of cesarean section for nonreassuring FHR tracing was 3.8% (30 cases with 16 cases in the AFI-monitored, and 14 in the 2!1 pocketemonitored groups, P = .608), and these cesarean sections comprised 43% of all sections in the AFI-monitored group (out of 37 cesarean deliveries) and 38% of all cesarean sections in the 2!1-monitored group (out of 37 c/s) (P = .637). Logistic regression analysis showed no difference between the groups with respect to the ability to identify patients who underwent cesarean section for nonreassuring FHR tracing during labor (P = .999). There were no differences between the groups for most of the neonatal outcomes (Table III). There was no difference in the rate of admission to the newborn ICU
664 Table II
Chauhan et al Peripartum outcomes stratified by sonographic assessment of AF volume, using either AFI or 2!1 pocket Group
Characteristic Oligohydramnios* Amnioinfusion Nonreassuring fetal heart rate (FHR) tracingy Nonreactive NST Variable decelerations Present Infl deliveryy Variability Min/absent Late decelerations Present Infl deliveryy Non-reassuring fetal heart rate (FHR) tracing Delivery mode Spont. Vaginal Assist. Vaginal Cesarean Reason for cesarean deliveryy Nonreassuring fetal heart rate (FHR) tracing Failure to progress Other Reason for assisted vaginal deliveryy,z Nonreassuring fetal heart rate (FHR) tracing Failure to progress Other
AFI (n = 530) 88 56 34 44
(17%) (11%) (63%) (8%)
2!1 (n = 558) 57 44 21 51
P value
(10%) (8%) (48%) (9%)
0.002 0.126 0.155 0.668
77 (15%) 67 (87%)
75 (13%) 60 (80%)
0.662 0.279
41 (8%)
63 (11%)
0.046
37 (7%) 33 (94%) 87 (16%)
27 (5%) 26 (96%) 83 (15%)
0.133 0.999 0.390
365 (69%) 128 (24%) 37 (7%)
405 (73%) 116 (21%) 37 (7%)
0.377
16 (43%) 7 (19%) 13 (35%)
14 (38%) 14 (38%) 9 (24%)
0.766
71 (56%) 44 (34%) 13 (10%)
69 (60%) 35 (30%) 11 (10%)
0.213
* AFI %5.0 cm in the AFI-monitored group, or 2!1 pocket not present in the 2!1 group, on the intrapartum AF assessment. y Percentages for subsets of patients with condition present are shown. One case of missing information about reason for cesarean delivery in AFI-monitored group, and one case of missing reason for assisted vaginal delivery in the 2!1 pocketemonitored group. z Three cases of nonreassuring fetal heart rate (FHR) tracing noted in spontaneous deliveries (2 in AFI-monitored group and 1 in the 2!1 pocket group).
(P = .328), but among the neonates admitted to the newborn ICU there was a higher proportion of admission for respiratory distress in the AFI-monitored group (P = .028). Among the preterm delivery, the rate of respiratory distress syndrome (RDS) was significantly higher among those randomized to the AFI (0.1%; 6/530) than to 2!1 pocket (0%; P = .009). These 6 premature deliveries were results of hypertensive disease (n = 2), preterm labor refractory to treatment with tocolytics (n = 3), and one case of fetal growth restriction and oligohydramnios (AFI of 4.0 cm) at 34 weeks. This last case was induced for abnormal growth and inadequate amniotic fluid, and had an uncomplicated intrapartum course with a vaginal delivery of a newborn weighing 1950 g, Apgar score of 5 at 5 minutes, and umbilical arterial pH of 7.20. Thus, the rate of iatrogenic prematurity due to the use of AFI was 2% (1/47; 95% CI 0.05e11). There were no cases of stillbirths in either group. The predictive accuracies of the 2 methods of assessing amniotic fluid are provided in Table IV. Both AFI and 2!1 monitoring methods were characterized by low
sensitivity to identify women likely to experience adverse outcomes with relatively high ability to predict normal outcomes. Use of likelihood ratios and guidelines provided by the Evidence-Based Medicine Working Group17 indicates that both methods of evaluating amniotic fluid are poor diagnostic tests in identifying patients that will undergo cesarean delivery for nonreassuring FHR tracing or deliver a newborn with depression or acidosis.
Comment The results of this randomized clinical trial are notable for 3 major findings. First, during modified BPP, use of the AFI rather than single deepest pocket to estimate the amniotic fluid volume increases the likelihood of intervention without a demonstrable benefit. High-risk patients were significantly more likely to be diagnosed as having oligohydramnios, be induced for an abnormal fluid volume, and undergo amnioinfusion if AFI is used for fetal assessment than the measurement of the largest
Chauhan et al Table III
665
Neonatal characteristics stratified by mode of monitoring, AFI and 2!1 pocket Group
Characteristic
AFI (n = 530)
2!1 (n = 558)
P value
Birth weight Birth weight distribution !1500 1500-2500 2500-4000 O4000 Meconium Present Thicky Umb artery pH !7.1* Apgar score !7 1 min 5 min Admission to NICU Admission reasonsy RDS/TTN Sepsis Other
3490 (3074-3856)
3250 (3105-3830)
0.780
1 25 414 90
(0.2%) (5%) (78%) (17%)
1 25 448 84
(0.2%) (5%) (80%) (15%)
0.843
65 39 14 68 12
(12%) (60%) (3%) (13%) (2%)
64 45 17 67 6
(12%) (70%) (3%) (12%) (1%)
0.685 0. 269 0.688 0.681 0.124
35 27 4 2
(7%) (82%) (12%) (6%)
29 15 11 1
(5%) (56%) (41%) (4%)
0.328 0.419 0.028
* There were 2 cases (ie, 0.4%) of pH !7.00 in each group. y Percentages for subsets of patients with condition present are shown.
Table IV
The predictive accuracies of adverse outcomes using amniotic fluid index or the presence/absence of a 2!1 pocket AFI (n = 530)
CD for distress AS !7 at 5 min UApH !7.10 UApH !7.00
2!1 (n = 558)
Sens
Spec
PPV
NPV
LR (C)
LR ()
Sens
Spec
PPV
NPV
LR (C)
LR ()
6.3% 8.3% 7.1% 50%
83.1% 83.2% 83.1% 83.5%
0.8% 0.6% 0.9% 0.3%
96.6% 97.5% 96.9% 99.8%
0.4 0.5 0.4 3.0
0.9 0.9 0.8 1.6
7.7% 0 5.9% 0
89.7% 89.7% 89.6% 89.7%
1.8% 0 1.8% 0
97.6% 98.8% 96.8% 99.6%
0.7 0 0.6 0
0.9 0.8 0.9 0.8
Sens, Sensitivity; Spec, specificity; PPV, positive predictive value; NPV, negative predictive value .
pocket. Despite the increased rate of intervention, 2 major outcomes of concern, cesarean delivery for nonreassuring FHR and neonatal acidosis, were similar in the 2 groups. Second, because of a false-positive test, iatrogenic prematurity does occur when modified biophysical is used for antepartum surveillance. While we noted its rate to be 2%, Miller et al17 reported that it was 1.5%. Though this complication was similar for the 2 groups, it may be due to the small sample size of this study. We did note a significantly higher rate of RDS among those who had assessment of amniotic fluid using the index rather than the single deepest pocket. The preterm deliveries in 5 of the 6 cases of RDS were secondary to medical (hypertension) or obstetric (spontaneous preterm labor) complications and use of AFI spuriously influenced the rate of RDS. Only one case of preterm delivery and RDS was secondary to the finding of AFI !5.0 cm. Thus, iatrogenic prematurity with associated complications, we think, occurs in 2% (1/47; 95% CI 0.05e11) of the cohorts that are managed with the use of
AFI. Randomized clinical trials with a larger sample size are needed to determine a more precise risk of iatrogenic RDS with use of AFI in the modified BPP. The third finding of this trial is that both methods of determining amniotic fluid are poor diagnostic tests in identifying peripartum complications (Table IV). Presently, there is a debate about the diagnostic ability of AFI to predict adverse outcomes. While Casey et al3 noted that with the use of AFI, antepartum detection of oligohydramnios is associated with increased perinatal morbidity and mortality, Magann et al6 considered AFI to be a poor diagnostic test. The limitation of these reports3,6 is that they were not randomized clinical trials, and the investigators did not report the likelihood ratio, a preferred way to evaluate a diagnostic test. The findings of this study are consistent with the previous report by Alfirevic et al,11 who randomized uncomplicated, post-term patients (gestational age of 290 days or more) to be monitored with computerized cardiotocography with AFI or maximum deepest pocket. These investigators considered the amniotic fluid to be
666 abnormal if the AFI was less than 7.3 cm or if the maximum deepest pocket was !1.8 cm, both of which are below the third percentile for 42 weeks. Despite the differences in the definition of inadequate amniotic fluid between our study and this published report,11 Alfirevic et al’s study also noted that significantly more post-term patients were diagnosed with oligohydramnios, and were subsequently induced if the AFI was used for antepartum surveillance. Both our randomized study and that of Alfirevic et al11 noted that the rate of cesarean delivery for nonreassuring FHR tracing and neonatal acidosis were similar, regardless of which technique was used to estimate the amniotic fluid volume. Amniotic fluid volume assessment is important in the surveillance of high-risk pregnancies. A low amniotic fluid volume is thought to represent inadequate uteroplacental perfusion with the shunting of blood to the fetal heart, adrenals, and brain with resultant poor renal perfusion and a reduced fetal urinary output. Before the investigations by Phelan et al4 suggesting that the AFI be used to evaluate amniotic fluid volume, the single deepest pocket was routinely used, and the definition of oligohydramnios was the absence of a 1!1 or 2!1 cm pocket.5 Surprisingly, without the benefit of a randomized clinical trial, the AFI was substituted for the single deepest pocket technique to estimate amniotic fluid volume. A possible reason for this substitution were the reports by Rutherford et al19 and Sarno et al,20 which noted that AFI %5.0 cm was significantly associated with an increased rate of cesarean delivery for nonreassuring FHR tracing and low Apgar scores. What was overlooked in this substitution was the subjective nature of these 2 end points. The incidence of neonatal acidosis, an objective assessment of fetal well-being, with AFI % vs O5.0 cm was not even addressed in the early reports.19,20 Subsequently, a meta-analysis of 42 reports over 10 years found only one study which correlated the rate of umbilical arterial pH !7.00 with AFI, and noted that neonatal acidosis occurs with similar frequency among high-risk patients with AFI %5.0 cm as those with O5.0 cm.10 More recent works by Magann et al6 and Casey et al3 have also not linked pH !7.00 with AFI O5.0 cm. If there is no association between pathologic acidosis and AFI, then continued use of this technique to assess amniotic fluid is unnecessary, especially because randomized clinical trials indicate an increase rate of intervention without improved outcome. The major shortcoming of this randomized trial is that the sample size was inadequate to evaluate meaningful, unambiguous outcomes like cerebral palsy or neonatal death. But considering the infrequent occurrence of these events, a randomized clinical trial to decrease their rate is a formidable undertaking. Assuming that umbilical arterial pH !7.00 occurs in 0.4%, then 8700 patients in each group are necessary to detect
Chauhan et al a 20% reduction in prevalence between the study arms (two-tailed alpha of 0.05 and power of 80%). It is noteworthy that Manning et al21 did report that with the use of the complete BPP they were able to significantly reduce the rate of cerebral palsy among high-risk patients. Specifically, they noted that the incidence of cerebral palsy among the general population (3.7 per 1000 live births) was significantly higher than the high-risk parturients who had BPP (1.3 per 1000 live births; P ! .0001). Admittedly, several criteria have been used to ensure adequate fluid during the BPPs,5 but none of them included use of AFI. Thus, if modified BPP is to be substituted for the 5 variables survey, then we should use one of the definitions used by Manning et al.21 In conclusion, use of single deepest pocket with modified BPP would decrease the rate of induction for suspected oligohydramnios. Unless randomized clinical trials show benefit of summing the measurements of 4 quadrants over dimensions of the deepest pocket, use of AFI during modified BPP should be avoided.
References 1. American College of Obstetricians and Gynecologists. Antenatal fetal surveillance. Washington (DC): American College of Obstetricians and Gynecologists; 1993. ACOG Practice Bulletin No. 9. 2. Golan A, Lin G, Evron S, Arieli S, Niv D, David MP. Oligohydramnios: maternal complications and fetal outcome in 145 cases. Gynecol Obstet Invest 1994;37:91-5. 3. Casey BM, McIntire DD, Bloom SL, Lucas MJ, Santos R, Twickler DM, et al. Pregnancy outcomes after antepartum diagnosis of oligohydramnios at or beyond 34 weeks’ gestation. Am J Obstet Gynecol 2000;182:909-12. 4. Phelan JP, Smith CV, Broussard P, Small M. Amniotic fluid volume assessment with the four-quadrant technique at 36-42 weeks’ gestation. J Repro Med 1987;32:540-2. 5. Magann EF, Isler CM, Chauhan SP, Martin JN Jr. Amniotic fluid volume estimation and the biophysical profile: a confusion of criteria. Obstet Gynecol 2000;96:640-2. 6. Magann EF, Chauhan SP, Kinsella MJ, McNamara MF, Whitworth NS, Morrison JC. Antenatal testing among 1001 patients at high risk: the role of ultrasonographic estimate of amniotic fluid volume. Am J Obstet Gynecol 1999;180:1330-6. 7. Rainford M, Adair R, Scialli A, Ghidini A, Spong CY. Amniotic fluid index in uncomplicated term pregnancy: prediction of outcome. J Reprod Med 2001;46:589-92. 8. Rutherford SE, Phelan JP, Smith CV, Jacobs N. The fourquadrant assessment of amniotic fluid volume: an adjunct to antepartum fetal heart rate testing. Obstet Gynecol 1987;70:353-6. 9. Voxman EG, Tran S, Wing DA. Low amniotic index as a predictor of adverse perinatal outcome. J Perinatol 2002;22:282-5. 10. Chauhan SP, Sanderson M, Hendrix NW, Magann EF, Devoe LD. Perinatal outcome and amniotic fluid index in the antepartum and intrapartum periods: a meta-analysis. Am J Obstet Gynecol 1999;181:1473-8. 11. Alfirevic Z, Luckas M, Walkinshaw SA, McFarlane M, Curran R. A randomized comparison between amniotic fluid index and maximum pool depth in the monitoring of post-term pregnancy. BJOG 1997;104:207-11. 12. Amano K, Saito K, Shoda T, Tani A, Youshihara H, Nishijima M. Elective induction of labor at 39 weeks of gestation: a prospective randomized trial. J Obstet Gynaecol Res 1999;25:33-7.
Chauhan et al 13. Nageotte MP, Towers CV, Asrat T, Freeman RK. Perinatal outcome with modified biophysical profile. Am J Obstet Gynecol 1994;170:1672-6. 14. 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 volume to perinatal outcome. Am J Obstet Gynecol 1984;150:245-9. 15. American College of Obstetricians and Gynecologists. Dystocia and the augmentation of labor. ACOG technical bulletin 218. Washington, DC: A; 1995. 16. American College of Obstetricians and Gynecologists. Fetal heart rate patterns: Monitoring, interpretation, and management. Washington, DC: American College of Obstetricians and Gynecologists; 1995. ACOG technical bulletin 207. 17. Jaeschke R, Guyatt GH, Sackett DL, for the Evidence-Based Medicine Working Group. Users’ guide to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients?. JAMA 1994;271:703-7. 18. Miller DA, Rabello YA, Paul RH. The modified biophysical profile: antepartum testing in the 1990s. Am J Obstet Gynecol 1996;174:82-7. 19. Rutherford SE, Phelan JP, Smith CV. The four-quadrant assessment of amniotic fluid volume: an adjunct to antepartum fetal heart rate testing. Obstet Gynecol 1987;70:353-6. 20. Sarno AP, Ahn MO, Brar HS, Phelan JP, Platt LD. Intrapartum Doppler velocimetry, amniotic fluid volume, and fetal heart rate as predictors of subsequent non-reassuring fetal heart rate (FHR) tracing: an initial report. Am J Obstet Gynecol 1989;161:1508-14. 21. Manning FA, Bondaji N, Harman CR, Casiro O, Menticoglou S, Morrison I, et al. Fetal assessment based on fetal biophysical profile scoring.VIII. Am J Obstet Gynecol 1998;178:696-706.
Discussion DR WILLIAM A. NEWTON, Tampa, Fla. Oligohydramnios has been demonstrated by a number of studies as an ominous prognostic sign of fetal well-being, especially in the high-risk and postdate pregnancy.1-3 There have been links between decreased amniotic fluid volume to stillbirth, fetal anomaly, abnormal fetal heart rate tracings in labor, increase in cesarean section for fetal distress, and possibly fetal acidosis.4 The concept of 4quadrant fluid measurement for amniotic fluid volume (AFV) was introduced in the mid 1980s. Since that time amniotic fluid index (AFI) has been widely accepted as an accurate means of defining oligohydramnios and its relationship to fetal compromise.2 However, in the past several years AFI has increasingly been questioned as the appropriate standard measurement to define oligohydramnios.5 Because many authors have noted that amniotic fluid volume alone may be a poor predictor of fetal wellbeing6 there has been the evolution of amniotic fluid volume incorporated into other diagnostic tests such as the biophysical profile and the modified biophysical profile. These fetal assessment tests greatly enhance safe management of the high-risk pregnancy but may also result in unwarranted intervention and morbidity.
667 Dr Chauhan’s study compares AFI versus measurement of a 2!1-cm single deepest pocket (SDP) incorporated into the modified biophysical profile as a predictor of fetal compromise. His results imply that AFI is not only less specific than SDP but more likely to result in a higher potential for unnecessary intervention through increased induction of labor for oligohydramnios. Using AFI with modified biophysical profile resulted in a 2% iatrogenic prematurity rate in this study. These findings seem contradictory to the 1987 study by Rutherford et al,7 which looked at 330 highrisk pregnancies comparing 4-quadrant AFV assessment and SDP. In that study AFI was found to have a high relationship to low Apgar score and cesarean section for fetal distress. Following Rutherford’s study the AFI was used as measurement of amniotic fluid volume for the modified biophysical profile. Coinciding with the acceptance of AFI as routine part of obstetric ultrasound, induction of labor for oligohydramnios increased from 26% in 1991 to 75% in 1996. In Dr Chauhan’s study, intrapartum oligohydramnios was diagnosed in 17% by AFI and only 10% by SDP. However, there was no significant difference in predictability of cesarean section for fetal distress, low Apgar score, or umbilical pH !7.1. Of note, clinicians managing labor were aware of whether oligohydramnios was determined by AFI or SDP. There was also the implication that AFI measurement was associated with increased potential morbidity, as noted by 13 more AFI patients undergoing amnioinfusion for presumed fetal distress. Admission to the NICU for transient tachypnea of the newborn and respiratory distress syndrome occurred in 12 more patients with AFI vs SDP, implying that in a larger sample size, AFI may be statistically associated with a higher iatrogenic morbidity than SDP. AFI as well as SDP are ‘‘poor diagnostic tests’’ for identifying peripartum complications. If used in the modified biophysical profile, the SDP should decrease the rate of induction of labor for suspected oligohydramnios. This study is impressive, but a much larger sample size is needed to look at the best method of determining oligohydramnios. For that matter, defining exactly what oligohydramnios is quantitatively and in what circumstance it reflects a poor prognostic sign is needed. The modified biophysical profile with assessment of AFV is an easy and convenient test for the evaluation of the high-risk patient. A reliable consensus standard of measurement is essential for a uniform diagnostic test. Dr Chauhan et al are congratulated on this interesting prospective clinical study. My questions for Dr Chauhan are: (1) in your study, the managing clinicians were aware of the method of determining oligohydramnios. Don’t you feel that this may have resulted in clinician bias in the management of individual cases? (2) What role, if any, do you feel that evolving technology