- Email: [email protected]
Transcervical intrapartum amnioinfusion for preterm premature rupture of the membranes
European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44 www.elsevier.com/locate/ejogrb
Transcervical intrapartum amnioinfusion for preterm premature rupture of the membranes Alberto Puertas *, Pilar Tirado, Isabel Pe´rez, Marı´a S. Lo´pez, Francisco Montoya, Jose´ M. Can˜izares, Jose´ A. Miranda Obstetrics and Gynaecology Service, Virgen de las Nieves University Hospital, Avenida Fuerzas Armadas s/n, 18014 Granada, Spain Received 28 August 2005; received in revised form 24 March 2006; accepted 20 April 2006
Abstract Objectives: To investigate the effect of transcervical amnioinfusion on the management of labour and neonatal outcomes in preterm premature rupture of the membranes. Study design: This clinical trial included 86 patients with premature rupture of the membranes between weeks 27 and 35 of gestation. Patients were randomly assigned to receive amnioinfusion via a two-way catheter or to the control group. Clinical management was otherwise the same in both groups. Results: Amnioinfusion decreased the frequency of variable decelerations in fetal heart rate (27.9% versus 53.5%, p < 0.05) and the rate of obstetric interventions motivated by nonreassuring fetal status (13.6% versus 52.4%, p < 0.05). At delivery, pH values were significantly higher in the treatment group than in the conventionally managed control group (median 7.29 versus 7.27). Conclusions: Intrapartum transcervical amnioinfusion for preterm premature rupture of the membranes reduced the number of interventions needed because of nonreassuring fetal status, and improved neonatal gasometric values without increasing maternal or fetal morbidity. # 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Amnioinfusion; Labour; Preterm premature rupture of the membranes
1. Introduction Transcervical amnioinfusion is a procedure that has been shown effective in increasing the volume of amniotic fluid during labour after spontaneous or artificial rupture of the membranes [1]. There is now adequate scientific evidence of its usefulness when the amniotic fluid is contaminated with meconium [2], when there is cardiotocographic evidence suggestive of umbilical cord compression [3], and in the presence of oligohydramnios as a result of term premature rupture of the membranes [1]. Amnioinfusion has also been proposed as a measure to improve fetal prognosis when preterm premature rupture of the membranes (pPROM) occurs [4]. By extension, the procedure may also be useful for any preterm labour that courses with oligohydramnios, * Corresponding author. Tel.: +34 958 02 00 89; fax: +34 958 02 02 26. E-mail addresses: [email protected], [email protected] (A. Puertas).
regardless of whether the cause of diminished amniotic fluid volume is rupture of the membranes or inadequate production. Premature deliveries are the cause of more than 60% of all neonatal morbidity and mortality [5]. The outcome is influenced most clearly by the choice of the route of delivery and by the measures used to achieve successful birth of the preterm fetus with as little physical or metabolic injury as possible [6]. In the light of these considerations, we designed the present study to evaluate the potential benefits of transcervical amnioinfusion during premature delivery with spontaneous rupture of the membranes.
2. Material and methods The study population consisted of pregnant women with preterm premature rupture of the membranes between weeks 27 and 35 (inclusive) of gestation, who were admitted to the
0301-2115/$ – see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2006.04.024
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
Virgen de la Nieves University Hospital in Granada (southern Spain) during the period from February 2001 to January 2003. Women with ruptured membranes were included in the study if labour had begun spontaneously or after induction. To ensure similarity among the women and fetuses in the study, we excluded patients with any of the following conditions: multiple gestation, presentation other than cephalic, cervical dilation >5 cm, cardiotocographic signs compatible with nonreassuring fetal status, meconiumstained amniotic fluid, umbilical cord prolapse, uterine scarring, placenta previa, premature detachment of the placenta, any vaginal bleeding of unknown cause, presence of oligohydramnios (amniotic fluid index <5) prior to premature rupture of the membranes, maternal infection that could be transmitted to the fetus (human immunodeficiency virus and Herpes simplex virus), fetal anomalies incompatible with life, or any known obstetric or maternal complication other than premature rupture of the membranes. The patients who fulfilled the inclusion criteria and signed an informed consent form for participation in the trial were allocated randomly to one of two groups with a procedure that used a table of random numbers and opaque sealed envelopes. Women in the study group received intrapartum transcervical amnioinfusion as described below, and the women in the control group were managed in the same way including insertion of an intrauterine pressure catheter, but without amnioinfusion. When the woman entered the dilation room, rupture of the membranes was verified and the cervix was examined to determine the Bishop index. All women who were Group B Streptococcus carriers received prophylactic intrapartum antibiotics. When indicated, oxytocin was infused intravenously at an initial rate of 2 mU/min, which was increased by 4 mU every 40 min until uterine activity was considered adequate. As soon as possible, a two-way catheter (Koala, HewlettPackard, Bo¨hlingen, Germany) was inserted in all women in both groups for direct, continuous recording of intrauterine pressure and for simultaneous amnioinfusion in the study group. After the catheter was inserted, we measured amniotic fluid index (AFI) with the four-quadrant technique [7], and considered this value as the baseline value. After 20 min, if no pathological alterations were seen in the basal cardiotocographic recording of fetal heart rate (FHR) and if no signs of uterine hyperactivity were noted, transcervical amnioinfusion was begun in women in the study group with physiological saline at 37 8C, at a rate of 600 ml/h during the first hour. After 1 h, AFI was determined, and if the index was greater than 15 amnioinfusion was stopped. In all other women in the study group amnioinfusion continued at a rate of 180 ml/h until the cervix was completely dilated. Fetal heart rate and uterine activity were recorded continuously throughout labour. Changes in FHR were
41
analysed with the Cabaniss classification [8] and were evaluated by an independent investigator. Changes in uterine activity (Alexandria units) and in basal uterine pressure (mmHg) were also analysed 1 h after the start of amnioinfusion. If uterine hypertonia was found (basal pressure >20 mmHg) the procedure was stopped and intrauterine fluid was withdrawn through the same catheter used for amnioinfusion if normal basal uterine pressure did not reappear within 10 min of suspending the infusion. The same procedure was used in the control group, although no liquid was infused into the uterus. In all newborns, we measured pH and blood gas concentrations in umbilical artery blood. Sample size was determined in a preliminary study of 16 cases and 16 controls with the same design as for the final study. The pH in umbilical artery blood was <7.20 in 36% of the newborns in the control group, and in 12% of the newborns in the study group. On the basis of these results, the sample size calculated to yield statistically significant results was 43 women in each arm, for an alpha error of 0.05 and a beta error of 0.10. Women included in the preliminary study were not included in the final study. All data were analysed with the Statigraphic Software Package System (V. 2.1). Student’s t-test was used to compare continuous variables, and the chi-squared test with Yates’ correction or Fisher’s exact test was used for qualitative or discrete variables. Normal distribution of continuous variables was verified with the Kolmogorov– Smirnoff test. The pH values were compared with Mann– Whitney’s nonparametric U test. Differences were considered significant when p < 0.05.
3. Results Comparison of the characteristics of the two groups in our study population showed the sample to be homogeneous (Table 1). In all women in the study group, premature rupture of the membranes was spontaneous, and the amniotic fluid was clear. Onset of labour was induced with oxytocin in 38 women (88.4%) in the study group, and in 33 women (76.7%) in the control group ( p = NS). Table 1 Characteristics of the study population
Number of women Gestational age (days) Primiparous Age (years) Epidural analgesia Group B Streptococcus Newborn body weight (g)
Amnioinfusion
No amnioinfusion
43 240 7 23 (53.5%) 30 7 38 (88.4%) 7 (16.3%) 2276 298
43 241 9 19 (44.2%) 28 5 32 (74.4%) 14 (32.6%) 2277 311
Results expressed as n (%) or mean S.D. p = not significant for all variables.
42
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
Table 2 Bishop index, duration of labour and changes in amniotic fluid index and basal uterine pressure
Initial Bishop index Duration of labour (min) Initial AFI AFI 60 min Increase in AFI Initial BUP (mmHg) BUP 60 min (mmHg) Increase in BUP
Amnioinfusion
No amnioinfusion
p
3.3 2.1 513 220 7.2 3.1 11.2 2.9 4.0 2.7 13.9 5.2 18.8 4.4 4.8 5.4
3.7 2.4 413 202 7.2 3.1 5.5 2.5 1.7 1.5 13.8 6.1 16.4 5.9 2.6 5.7
NS 0.03 NS <0.01 <0.01 NS 0.03 NS
AFI: amniotic fluid index; BUP: basal uterine pressure. Table 3 Outcome of labour
Spontaneous delivery Operative delivery Operative vaginal delivery Assisted expulsion NRFS Caesarian delivery Failure of induction NRFS Lack of progression Intervention because of NRFS
Amnioinfusion
No amnioinfusion
p
21/43 (48.8%) 22/43 (51.2%) 12/43 (27.9%)
22/43 (51.2%) 21/43 (48.8%) 10/43 (23.3%)
NS NS NS
12/12 (100%) 0 10/43 (23.3%) 6/10 (60.0%) 3/10 (30.0%) 1/10 (10.0%) 3/22 (13.6%)
6/10 (60.0%) 4/10 (40.0%) 11/43 (25.6%) 4/11 (36.4%) 7/11 (63.6%) 0 11/21 (52.4%)
0.03 0.06 NS NS NS NS 0.01
Table 2 shows the changes in Bishop index, progression of labour, uterine pressure and AFI in the two groups. The duration of labour was longer, and the volume of amniotic fluid after 60 min was greater, in the amnioinfusion group. Mean duration of amnioinfusion was 254 143 min, and the most frequent cause of suspending the procedure was complete dilation of the cervix (72.1%), followed by caesarian delivery (13.9%). In five women, basal uterine Table 4 Fetal heart rate variations in the two groups
Results expressed as n (%).
Amnioinfusion Days of puerperal hospital stay Puerperal morbidity Puerperal infectious morbidity Maximum intrapartum temperature (8C) Days of neonatal hospital stay Neonatal morbidity Neonatal infectious morbidity Umbilical cord abnormality
No amnioinfusion
3.6 1.8
3.6 1.9
6 (14.0%) 0
5 (11.6%) 1 (2.3%)
36.8 0.5
36.7 0.5
12 9.1
11 9.9
5 (11.6%) 1 (2.3%)
10 (23.2%) 1 (2.32%)
8 (18.6%)
8 (18.6%)
Results expressed as n (%) or mean S.D. p = not significant for all variables. Table 6 Neonatal blood gas values in the umbilical artery Amnioinfusion pH (median) pO2 (mmHg) pCO2 (mmHg) CO3H (mM/l) Base deficit (mM/l) Arterial pH 7.20
7.29 18 5.14 47.38 6.64 22.02 1.96 3.71 1.95 2 (4.65%)
No amnioinfusion
p
7.27 15.16 5.64 54.75 11.74 21.53 2.56 4.8 2.49 8 (16.8%)
0.043 0.016 <0.001 NS 0.027 0.092
Except for pH values, the results are expressed as the mean S.D. or as n (%). p = not significant for all variables.
Results expressed as n (%); NRFS: nonreassuring fetal status.
Early decelerations Late decelerations Late decelerations with absent variability Variable decelerations Mild Moderate Severe Typical Atypical Delayed variable decelerations Prolonged decelerations Altered baseline Altered variability Lack of reactivity
Table 5 Puerperal and neonatal morbidity
Amnioinfusion
No amnioinfusion
p
6 (14%) 3 (7%) 1 (2.3%)
8 (18.6%) 5 (11.7%) 2 (4.6%)
NS NS NS
12/43 (27.9%) 5 (11.6%) 6 (13.9%) 1 (2.3%) 8 (18.6%) 4 (9.3%) 0
23/43 (53.5%) 12 (27%) 8 (18.6%) 3 (6.9%) 11 (25.6%) 12 (27.9%) 3 (7%)
0.03 NS NS NS NS 0.05 NS
0 1 (2.3%) 7 (16.3%) 2 (4.6%)
1 (2.3%) 1 (2.3%) 11 (25.6%) 4 (9.3%)
NS NS NS NS
pressure increased to more than 20 mmHg, and in one woman, AFI surpassed 15; the procedure was suspended in both patients. The outcome of labour and type of delivery are shown in Table 3. The patterns of change in FHR (Table 4) showed that in the amnioinfusion group, variable decelerations were generally less frequent because of the lower frequency of atypical decelerations in this group. Table 5 shows the results for puerperal morbidity and mortality; no significant differences were seen between the two groups for any of the variables analysed here. Fetal blood gas concentrations in the umbilical artery (O2, pCO2 and base deficit) are shown in Table 6. The values were significantly more favourable in the amnioinfusion group, although the difference between groups in the number of newborns with pH less than 7.20 was not significant ( p = 0.092).
4. Discussion The homogeneity of the populations compared in this clinical trial, and the absence of obstetric disorders that might cause neonatal anomalies other than those related with prematurity, make our study ideal to analyse the usefulness of intrapartum transcervical amnioinfusion in women with
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
premature rupture of the membranes. Previous studies that examined the use of amnioinfusion for this indication concluded that the scientific evidence in favour of amnioinfusion was inadequate [9]. This procedure has been shown to be of benefit in labour characterized by variable decelerations in FHR, meconium-stained amniotic fluid, and oligohydramnios [1,2,10,11]. The aim of amnioinfusion is to restore physiological conditions by ensuring a sufficient volume of intrauterine fluid without increasing basal uterine tone to values considered pathological. One hour after amnioinfusion was begun, recovery of amniotic fluid volume in the study group confirmed the hypothesis that this technique is based on. In the control group, however, we found that amniotic fluid was lost. Despite the increase in basal uterine pressure seen 60 min after the start of infusion, pressure remained within normal limits. Like other authors [12], we believe that periodic measurements of AFI are not needed as long as intrauterine pressure is appropriately monitored. Indeed, the reason for stopping amnioinfusion was completion of dilation in more than 70% of the cases. In women for whom we stopped amnioinfusion because basal uterine pressure increased to more than 20 mmHg, normal values reappeared once the procedure was stopped. In view of the fact that the women in this study had normal pregnancies until premature rupture of the membranes, it was to be expected that there were no significant differences between the groups in the alterations in FHR that indicated placental compromise, or in alterations caused by maternal hypotension. As noted by Nageotte et al. [4], the most frequent alteration in FHR pattern in the population of women with preterm premature rupture of the membranes was repetitive variable decelerations, which may be related with umbilical cord compression secondary to insufficient amniotic fluid. When the volume of intrauterine fluid increases, the likelihood of umbilical cord and placental compression decreases. In the present study, this improvement in the fetal environment was translated as a statistically significant decrease in atypical variable decelerations in the amnioinfusion group. Similar results were reported in studies of preterm [4] and term labour [1]. In this connection, our results are also consistent with the findings of a systematic review by Hofmeyr [3] on the use of amnioinfusion when labour is characterized by cardiotocographic signs associated with umbilical cord compression. This author concluded that amnioinfusion was able to significantly reduce FHR decelerations and the rate of caesarian deliveries performed because of nonreassuring fetal status. The decrease in cardiotocographic alterations may be directly responsible for the significantly lower rate of obstetric interventions indicated for nonreassuring fetal status and the better acid–base balance based on gasometric values in umbilical cord artery blood at birth. As in the study by Nageotte et al. [4], we found a nonsignificant decrease in
43
the rate of caesarian deliveries motivated by nonreassuring fetal status. In the amnioinfusion group, more than half of the caesarian deliveries were motivated by failure of induction (60%), and only 30% were motivated by nonreassuring fetal status, versus the 63% rate of caesarian deliveries in the control group for this indication. In the review by Hofmeyr cited earlier [3], a statistically significant decrease was also found in the numbers of caesarian deliveries performed because of suspected fetal distress. In our amnioinfusion group, the indication for operative vaginal delivery (12/43) was to shorten the expulsion phase in all cases. In contrast, 40% of the operative vaginal deliveries in the control group were motivated by nonreassuring fetal status. Unsurprisingly, in view of the differing circumstances surrounding labour in the two groups we compared, fewer infants were born with a pH of <7.20 in the amnioinfusion group; the difference with the control group approached statistical significance at p = 0.09. In the study group, moreover, gasometric values were significantly better in general. These results are consistent with the mechanism of action of amnioinfusion: by decreasing the interruptions in blood flow through the umbilical cord, the procedure facilitates oxygen and carbon dioxide exchange. The improvements in cardiotocographic results, the significant decrease in the number of surgical procedures motivated by nonreassuring fetal status, and the improved acid–base balance in newborns in the amnioinfusion group all suggest that these differences were related with restoration of an adequate volume of amniotic fluid in women with preterm premature rupture of the membranes. We believe our results, along with those of those of earlier studies, show that amnioinfusion is relatively simple to perform, and well accepted by women. These features make amnioinfusion a useful procedure for the management of preterm labour and delivery.
Acknowledgements This study was supported by Servicio Andaluz de Salud Research Project 209/00. We thank K. Shashok for translating the original manuscript into English.
References [1] Min˜o M, Puertas A, Miranda JA, Herruzo AJ. Amnioinfusion in term labour with low amniotic fluid due to rupture of membranes: a new indication. Eur J Obstet Gynecol Reprod Biol 1999;82:29–34. [2] Hofmeyr GJ. Amnioinfusion for meconium-stained liquor in labour (Cochrane Review). The Cochrane Library, Issue 4: Oxford; Update Software, 1999. [3] Hofmeyr GJ. Amnioinfusion for umbilical cord compression in labour (Cochrane Review). The Cochrane Library, Issue 4: Oxford; Update Software, 1999.
44
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
[4] Nageotte MP, Freeman RK, Garite TJ, Dorchester W. Prophylactic intrapartum amnioinfusion in patients with premature rupture of membranes. Am J Obstet Gynecol 1985;153:557–62. [5] Cabero L. Parto petre´rmino. In: Cabero L, editor. Riesgo elevado obste´trico. Barcelona: Masson; 1996. p. 47–95. [6] Puertas A. Asistencia al nacimiento del feto prematuro. In: Hospital Universitario ‘‘Virgen de las Nieves’’, ed. Actualizacio´n en Obstetricia y Ginecologı´a. Granada: Reprografı´a Digital; 2003. p. 65–82. [7] Phelan JP, Smith CV, Broussard P, Small M. Amniotic fluid volume assessment with the four-quadrant technique at 36–42 week gestation. J Reprod Med 1987;32:540–2. [8] Cabaniss ML, editor. Monitorizacio´n fetal electro´nica. Interpretacio´n. Barcelona: Masson; 1995.
[9] Hofmeyr GJ. Amnioinfusion for preterm rupture of membranes (Cochrane Review). The Cochrane Library, Issue 1: Oxford; Update Software, 2003. [10] Miyazaki FS, Nevarez F. Saline amnioinfusion for relief of repetitive variable decelerations: a prospective randomized study. Am J Obstet Gynecol 1985;153:301–6. [11] Pitt C, Sanchez-Ramos L, Kaunitz AM, Gaudier F. Prophylactic amnioinfusion for intrapartum oligohydramnios: a metaanalysis of randomized controlled trials. Obstet Gynecol 2000;96: 861–6. [12] Strong TH, Hetzler G, Sarno AP, Paul RH. Prophylactic intrapartum amnioinfusion: a randomized clinical trial. Am J Obstet Gyneocol 1990;162:1370–5.
Transcervical intrapartum amnioinfusion for preterm premature rupture of the membranes Alberto Puertas *, Pilar Tirado, Isabel Pe´rez, Marı´a S. Lo´pez, Francisco Montoya, Jose´ M. Can˜izares, Jose´ A. Miranda Obstetrics and Gynaecology Service, Virgen de las Nieves University Hospital, Avenida Fuerzas Armadas s/n, 18014 Granada, Spain Received 28 August 2005; received in revised form 24 March 2006; accepted 20 April 2006
Abstract Objectives: To investigate the effect of transcervical amnioinfusion on the management of labour and neonatal outcomes in preterm premature rupture of the membranes. Study design: This clinical trial included 86 patients with premature rupture of the membranes between weeks 27 and 35 of gestation. Patients were randomly assigned to receive amnioinfusion via a two-way catheter or to the control group. Clinical management was otherwise the same in both groups. Results: Amnioinfusion decreased the frequency of variable decelerations in fetal heart rate (27.9% versus 53.5%, p < 0.05) and the rate of obstetric interventions motivated by nonreassuring fetal status (13.6% versus 52.4%, p < 0.05). At delivery, pH values were significantly higher in the treatment group than in the conventionally managed control group (median 7.29 versus 7.27). Conclusions: Intrapartum transcervical amnioinfusion for preterm premature rupture of the membranes reduced the number of interventions needed because of nonreassuring fetal status, and improved neonatal gasometric values without increasing maternal or fetal morbidity. # 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Amnioinfusion; Labour; Preterm premature rupture of the membranes
1. Introduction Transcervical amnioinfusion is a procedure that has been shown effective in increasing the volume of amniotic fluid during labour after spontaneous or artificial rupture of the membranes [1]. There is now adequate scientific evidence of its usefulness when the amniotic fluid is contaminated with meconium [2], when there is cardiotocographic evidence suggestive of umbilical cord compression [3], and in the presence of oligohydramnios as a result of term premature rupture of the membranes [1]. Amnioinfusion has also been proposed as a measure to improve fetal prognosis when preterm premature rupture of the membranes (pPROM) occurs [4]. By extension, the procedure may also be useful for any preterm labour that courses with oligohydramnios, * Corresponding author. Tel.: +34 958 02 00 89; fax: +34 958 02 02 26. E-mail addresses: [email protected], [email protected] (A. Puertas).
regardless of whether the cause of diminished amniotic fluid volume is rupture of the membranes or inadequate production. Premature deliveries are the cause of more than 60% of all neonatal morbidity and mortality [5]. The outcome is influenced most clearly by the choice of the route of delivery and by the measures used to achieve successful birth of the preterm fetus with as little physical or metabolic injury as possible [6]. In the light of these considerations, we designed the present study to evaluate the potential benefits of transcervical amnioinfusion during premature delivery with spontaneous rupture of the membranes.
2. Material and methods The study population consisted of pregnant women with preterm premature rupture of the membranes between weeks 27 and 35 (inclusive) of gestation, who were admitted to the
0301-2115/$ – see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2006.04.024
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
Virgen de la Nieves University Hospital in Granada (southern Spain) during the period from February 2001 to January 2003. Women with ruptured membranes were included in the study if labour had begun spontaneously or after induction. To ensure similarity among the women and fetuses in the study, we excluded patients with any of the following conditions: multiple gestation, presentation other than cephalic, cervical dilation >5 cm, cardiotocographic signs compatible with nonreassuring fetal status, meconiumstained amniotic fluid, umbilical cord prolapse, uterine scarring, placenta previa, premature detachment of the placenta, any vaginal bleeding of unknown cause, presence of oligohydramnios (amniotic fluid index <5) prior to premature rupture of the membranes, maternal infection that could be transmitted to the fetus (human immunodeficiency virus and Herpes simplex virus), fetal anomalies incompatible with life, or any known obstetric or maternal complication other than premature rupture of the membranes. The patients who fulfilled the inclusion criteria and signed an informed consent form for participation in the trial were allocated randomly to one of two groups with a procedure that used a table of random numbers and opaque sealed envelopes. Women in the study group received intrapartum transcervical amnioinfusion as described below, and the women in the control group were managed in the same way including insertion of an intrauterine pressure catheter, but without amnioinfusion. When the woman entered the dilation room, rupture of the membranes was verified and the cervix was examined to determine the Bishop index. All women who were Group B Streptococcus carriers received prophylactic intrapartum antibiotics. When indicated, oxytocin was infused intravenously at an initial rate of 2 mU/min, which was increased by 4 mU every 40 min until uterine activity was considered adequate. As soon as possible, a two-way catheter (Koala, HewlettPackard, Bo¨hlingen, Germany) was inserted in all women in both groups for direct, continuous recording of intrauterine pressure and for simultaneous amnioinfusion in the study group. After the catheter was inserted, we measured amniotic fluid index (AFI) with the four-quadrant technique [7], and considered this value as the baseline value. After 20 min, if no pathological alterations were seen in the basal cardiotocographic recording of fetal heart rate (FHR) and if no signs of uterine hyperactivity were noted, transcervical amnioinfusion was begun in women in the study group with physiological saline at 37 8C, at a rate of 600 ml/h during the first hour. After 1 h, AFI was determined, and if the index was greater than 15 amnioinfusion was stopped. In all other women in the study group amnioinfusion continued at a rate of 180 ml/h until the cervix was completely dilated. Fetal heart rate and uterine activity were recorded continuously throughout labour. Changes in FHR were
41
analysed with the Cabaniss classification [8] and were evaluated by an independent investigator. Changes in uterine activity (Alexandria units) and in basal uterine pressure (mmHg) were also analysed 1 h after the start of amnioinfusion. If uterine hypertonia was found (basal pressure >20 mmHg) the procedure was stopped and intrauterine fluid was withdrawn through the same catheter used for amnioinfusion if normal basal uterine pressure did not reappear within 10 min of suspending the infusion. The same procedure was used in the control group, although no liquid was infused into the uterus. In all newborns, we measured pH and blood gas concentrations in umbilical artery blood. Sample size was determined in a preliminary study of 16 cases and 16 controls with the same design as for the final study. The pH in umbilical artery blood was <7.20 in 36% of the newborns in the control group, and in 12% of the newborns in the study group. On the basis of these results, the sample size calculated to yield statistically significant results was 43 women in each arm, for an alpha error of 0.05 and a beta error of 0.10. Women included in the preliminary study were not included in the final study. All data were analysed with the Statigraphic Software Package System (V. 2.1). Student’s t-test was used to compare continuous variables, and the chi-squared test with Yates’ correction or Fisher’s exact test was used for qualitative or discrete variables. Normal distribution of continuous variables was verified with the Kolmogorov– Smirnoff test. The pH values were compared with Mann– Whitney’s nonparametric U test. Differences were considered significant when p < 0.05.
3. Results Comparison of the characteristics of the two groups in our study population showed the sample to be homogeneous (Table 1). In all women in the study group, premature rupture of the membranes was spontaneous, and the amniotic fluid was clear. Onset of labour was induced with oxytocin in 38 women (88.4%) in the study group, and in 33 women (76.7%) in the control group ( p = NS). Table 1 Characteristics of the study population
Number of women Gestational age (days) Primiparous Age (years) Epidural analgesia Group B Streptococcus Newborn body weight (g)
Amnioinfusion
No amnioinfusion
43 240 7 23 (53.5%) 30 7 38 (88.4%) 7 (16.3%) 2276 298
43 241 9 19 (44.2%) 28 5 32 (74.4%) 14 (32.6%) 2277 311
Results expressed as n (%) or mean S.D. p = not significant for all variables.
42
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
Table 2 Bishop index, duration of labour and changes in amniotic fluid index and basal uterine pressure
Initial Bishop index Duration of labour (min) Initial AFI AFI 60 min Increase in AFI Initial BUP (mmHg) BUP 60 min (mmHg) Increase in BUP
Amnioinfusion
No amnioinfusion
p
3.3 2.1 513 220 7.2 3.1 11.2 2.9 4.0 2.7 13.9 5.2 18.8 4.4 4.8 5.4
3.7 2.4 413 202 7.2 3.1 5.5 2.5 1.7 1.5 13.8 6.1 16.4 5.9 2.6 5.7
NS 0.03 NS <0.01 <0.01 NS 0.03 NS
AFI: amniotic fluid index; BUP: basal uterine pressure. Table 3 Outcome of labour
Spontaneous delivery Operative delivery Operative vaginal delivery Assisted expulsion NRFS Caesarian delivery Failure of induction NRFS Lack of progression Intervention because of NRFS
Amnioinfusion
No amnioinfusion
p
21/43 (48.8%) 22/43 (51.2%) 12/43 (27.9%)
22/43 (51.2%) 21/43 (48.8%) 10/43 (23.3%)
NS NS NS
12/12 (100%) 0 10/43 (23.3%) 6/10 (60.0%) 3/10 (30.0%) 1/10 (10.0%) 3/22 (13.6%)
6/10 (60.0%) 4/10 (40.0%) 11/43 (25.6%) 4/11 (36.4%) 7/11 (63.6%) 0 11/21 (52.4%)
0.03 0.06 NS NS NS NS 0.01
Table 2 shows the changes in Bishop index, progression of labour, uterine pressure and AFI in the two groups. The duration of labour was longer, and the volume of amniotic fluid after 60 min was greater, in the amnioinfusion group. Mean duration of amnioinfusion was 254 143 min, and the most frequent cause of suspending the procedure was complete dilation of the cervix (72.1%), followed by caesarian delivery (13.9%). In five women, basal uterine Table 4 Fetal heart rate variations in the two groups
Results expressed as n (%).
Amnioinfusion Days of puerperal hospital stay Puerperal morbidity Puerperal infectious morbidity Maximum intrapartum temperature (8C) Days of neonatal hospital stay Neonatal morbidity Neonatal infectious morbidity Umbilical cord abnormality
No amnioinfusion
3.6 1.8
3.6 1.9
6 (14.0%) 0
5 (11.6%) 1 (2.3%)
36.8 0.5
36.7 0.5
12 9.1
11 9.9
5 (11.6%) 1 (2.3%)
10 (23.2%) 1 (2.32%)
8 (18.6%)
8 (18.6%)
Results expressed as n (%) or mean S.D. p = not significant for all variables. Table 6 Neonatal blood gas values in the umbilical artery Amnioinfusion pH (median) pO2 (mmHg) pCO2 (mmHg) CO3H (mM/l) Base deficit (mM/l) Arterial pH 7.20
7.29 18 5.14 47.38 6.64 22.02 1.96 3.71 1.95 2 (4.65%)
No amnioinfusion
p
7.27 15.16 5.64 54.75 11.74 21.53 2.56 4.8 2.49 8 (16.8%)
0.043 0.016 <0.001 NS 0.027 0.092
Except for pH values, the results are expressed as the mean S.D. or as n (%). p = not significant for all variables.
Results expressed as n (%); NRFS: nonreassuring fetal status.
Early decelerations Late decelerations Late decelerations with absent variability Variable decelerations Mild Moderate Severe Typical Atypical Delayed variable decelerations Prolonged decelerations Altered baseline Altered variability Lack of reactivity
Table 5 Puerperal and neonatal morbidity
Amnioinfusion
No amnioinfusion
p
6 (14%) 3 (7%) 1 (2.3%)
8 (18.6%) 5 (11.7%) 2 (4.6%)
NS NS NS
12/43 (27.9%) 5 (11.6%) 6 (13.9%) 1 (2.3%) 8 (18.6%) 4 (9.3%) 0
23/43 (53.5%) 12 (27%) 8 (18.6%) 3 (6.9%) 11 (25.6%) 12 (27.9%) 3 (7%)
0.03 NS NS NS NS 0.05 NS
0 1 (2.3%) 7 (16.3%) 2 (4.6%)
1 (2.3%) 1 (2.3%) 11 (25.6%) 4 (9.3%)
NS NS NS NS
pressure increased to more than 20 mmHg, and in one woman, AFI surpassed 15; the procedure was suspended in both patients. The outcome of labour and type of delivery are shown in Table 3. The patterns of change in FHR (Table 4) showed that in the amnioinfusion group, variable decelerations were generally less frequent because of the lower frequency of atypical decelerations in this group. Table 5 shows the results for puerperal morbidity and mortality; no significant differences were seen between the two groups for any of the variables analysed here. Fetal blood gas concentrations in the umbilical artery (O2, pCO2 and base deficit) are shown in Table 6. The values were significantly more favourable in the amnioinfusion group, although the difference between groups in the number of newborns with pH less than 7.20 was not significant ( p = 0.092).
4. Discussion The homogeneity of the populations compared in this clinical trial, and the absence of obstetric disorders that might cause neonatal anomalies other than those related with prematurity, make our study ideal to analyse the usefulness of intrapartum transcervical amnioinfusion in women with
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
premature rupture of the membranes. Previous studies that examined the use of amnioinfusion for this indication concluded that the scientific evidence in favour of amnioinfusion was inadequate [9]. This procedure has been shown to be of benefit in labour characterized by variable decelerations in FHR, meconium-stained amniotic fluid, and oligohydramnios [1,2,10,11]. The aim of amnioinfusion is to restore physiological conditions by ensuring a sufficient volume of intrauterine fluid without increasing basal uterine tone to values considered pathological. One hour after amnioinfusion was begun, recovery of amniotic fluid volume in the study group confirmed the hypothesis that this technique is based on. In the control group, however, we found that amniotic fluid was lost. Despite the increase in basal uterine pressure seen 60 min after the start of infusion, pressure remained within normal limits. Like other authors [12], we believe that periodic measurements of AFI are not needed as long as intrauterine pressure is appropriately monitored. Indeed, the reason for stopping amnioinfusion was completion of dilation in more than 70% of the cases. In women for whom we stopped amnioinfusion because basal uterine pressure increased to more than 20 mmHg, normal values reappeared once the procedure was stopped. In view of the fact that the women in this study had normal pregnancies until premature rupture of the membranes, it was to be expected that there were no significant differences between the groups in the alterations in FHR that indicated placental compromise, or in alterations caused by maternal hypotension. As noted by Nageotte et al. [4], the most frequent alteration in FHR pattern in the population of women with preterm premature rupture of the membranes was repetitive variable decelerations, which may be related with umbilical cord compression secondary to insufficient amniotic fluid. When the volume of intrauterine fluid increases, the likelihood of umbilical cord and placental compression decreases. In the present study, this improvement in the fetal environment was translated as a statistically significant decrease in atypical variable decelerations in the amnioinfusion group. Similar results were reported in studies of preterm [4] and term labour [1]. In this connection, our results are also consistent with the findings of a systematic review by Hofmeyr [3] on the use of amnioinfusion when labour is characterized by cardiotocographic signs associated with umbilical cord compression. This author concluded that amnioinfusion was able to significantly reduce FHR decelerations and the rate of caesarian deliveries performed because of nonreassuring fetal status. The decrease in cardiotocographic alterations may be directly responsible for the significantly lower rate of obstetric interventions indicated for nonreassuring fetal status and the better acid–base balance based on gasometric values in umbilical cord artery blood at birth. As in the study by Nageotte et al. [4], we found a nonsignificant decrease in
43
the rate of caesarian deliveries motivated by nonreassuring fetal status. In the amnioinfusion group, more than half of the caesarian deliveries were motivated by failure of induction (60%), and only 30% were motivated by nonreassuring fetal status, versus the 63% rate of caesarian deliveries in the control group for this indication. In the review by Hofmeyr cited earlier [3], a statistically significant decrease was also found in the numbers of caesarian deliveries performed because of suspected fetal distress. In our amnioinfusion group, the indication for operative vaginal delivery (12/43) was to shorten the expulsion phase in all cases. In contrast, 40% of the operative vaginal deliveries in the control group were motivated by nonreassuring fetal status. Unsurprisingly, in view of the differing circumstances surrounding labour in the two groups we compared, fewer infants were born with a pH of <7.20 in the amnioinfusion group; the difference with the control group approached statistical significance at p = 0.09. In the study group, moreover, gasometric values were significantly better in general. These results are consistent with the mechanism of action of amnioinfusion: by decreasing the interruptions in blood flow through the umbilical cord, the procedure facilitates oxygen and carbon dioxide exchange. The improvements in cardiotocographic results, the significant decrease in the number of surgical procedures motivated by nonreassuring fetal status, and the improved acid–base balance in newborns in the amnioinfusion group all suggest that these differences were related with restoration of an adequate volume of amniotic fluid in women with preterm premature rupture of the membranes. We believe our results, along with those of those of earlier studies, show that amnioinfusion is relatively simple to perform, and well accepted by women. These features make amnioinfusion a useful procedure for the management of preterm labour and delivery.
Acknowledgements This study was supported by Servicio Andaluz de Salud Research Project 209/00. We thank K. Shashok for translating the original manuscript into English.
References [1] Min˜o M, Puertas A, Miranda JA, Herruzo AJ. Amnioinfusion in term labour with low amniotic fluid due to rupture of membranes: a new indication. Eur J Obstet Gynecol Reprod Biol 1999;82:29–34. [2] Hofmeyr GJ. Amnioinfusion for meconium-stained liquor in labour (Cochrane Review). The Cochrane Library, Issue 4: Oxford; Update Software, 1999. [3] Hofmeyr GJ. Amnioinfusion for umbilical cord compression in labour (Cochrane Review). The Cochrane Library, Issue 4: Oxford; Update Software, 1999.
44
A. Puertas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 131 (2007) 40–44
[4] Nageotte MP, Freeman RK, Garite TJ, Dorchester W. Prophylactic intrapartum amnioinfusion in patients with premature rupture of membranes. Am J Obstet Gynecol 1985;153:557–62. [5] Cabero L. Parto petre´rmino. In: Cabero L, editor. Riesgo elevado obste´trico. Barcelona: Masson; 1996. p. 47–95. [6] Puertas A. Asistencia al nacimiento del feto prematuro. In: Hospital Universitario ‘‘Virgen de las Nieves’’, ed. Actualizacio´n en Obstetricia y Ginecologı´a. Granada: Reprografı´a Digital; 2003. p. 65–82. [7] Phelan JP, Smith CV, Broussard P, Small M. Amniotic fluid volume assessment with the four-quadrant technique at 36–42 week gestation. J Reprod Med 1987;32:540–2. [8] Cabaniss ML, editor. Monitorizacio´n fetal electro´nica. Interpretacio´n. Barcelona: Masson; 1995.
[9] Hofmeyr GJ. Amnioinfusion for preterm rupture of membranes (Cochrane Review). The Cochrane Library, Issue 1: Oxford; Update Software, 2003. [10] Miyazaki FS, Nevarez F. Saline amnioinfusion for relief of repetitive variable decelerations: a prospective randomized study. Am J Obstet Gynecol 1985;153:301–6. [11] Pitt C, Sanchez-Ramos L, Kaunitz AM, Gaudier F. Prophylactic amnioinfusion for intrapartum oligohydramnios: a metaanalysis of randomized controlled trials. Obstet Gynecol 2000;96: 861–6. [12] Strong TH, Hetzler G, Sarno AP, Paul RH. Prophylactic intrapartum amnioinfusion: a randomized clinical trial. Am J Obstet Gyneocol 1990;162:1370–5.