Objective
We sought to determine the effectiveness of continuous fetal monitoring in monoamniotic twins and assess the percentage of successful monitoring across gestation.
Study Design
This was a single-center retrospective cohort analysis of monoamniotic twins. Each subject’s entire electronic fetal heart tracing was reviewed. The primary outcome measure was the percentage of time that 0, 1, and 2 fetuses were successfully monitored. To compare the effectiveness of monitoring across gestation, these data were stratified by gestational age (<27, 27-30, and >30 weeks).
Results
A total of 10,402 hours of fetal monitoring were reviewed for 17 monoamniotic twin pairs. Successful monitoring of 0, 1, and 2 fetuses occurred 21.2%, 27.1%, and 51.6% of the time, respectively. Successful monitoring of both fetuses was positively correlated with gestational age (<27 weeks 37%; 27-30 weeks 51%; >30 weeks 57%; P < .007).
Conclusion
Successful monitoring of both fetuses occurred 51.6% of the time. Continuous monitoring efficiency improved with advancing gestational age.
Monoamniotic twinning is a rare occurrence accounting for 1% of all monozygotic gestations but is associated with a disproportionately high perinatal mortality rate. Older literature suggests perinatal mortality rates up to 70%; however, with improvements in antenatal diagnosis, fetal monitoring, and neonatal care, more recent data suggest perinatal mortality rates of 10-20%.
Monoamniotic twins carry the same risks as other monochorionic multifetal gestations including preterm delivery, twin-twin transfusion syndrome, and congenital malformations. However, cord entanglement is unique to monoamniotic and pseudomonoamniotic twin gestations and accounts for >50% of fetal deaths in these cases. Pseudomonoamniotic twins are monochorionic diamniotic gestations in which the preexisting intertwin membrane has ruptured either spontaneously or iatrogenically from amniocentesis or septostomy at the time of an invasive procedure. Once this membrane is disrupted, these gestations are functionally similar to monoamniotic twin gestations, with a significant risk of developing cord entanglement and an increased perinatal mortality rate.
Accurate prenatal diagnosis, intensive fetal surveillance, and a timed cesarean delivery can improve the perinatal mortality rate of monoamniotic and pseudomonoamniotic twins. However, controversy exists over the relative value of inpatient vs outpatient management, degree of antenatal monitoring, and timing of delivery necessary to optimize perinatal outcomes.
Previous studies have shown inpatient management of monoamniotic twins decreases the perinatal mortality rate. These studies showed that elective hospitalization for fetal monitoring resulted in decreased fetal death rates, increased birthweights, and decreased neonatal morbidity. However, there is no consensus on the optimal level of fetal monitoring needed to achieve this reduction in perinatal morbidity and mortality.
The main goal of intensive inpatient fetal heart rate monitoring is to prevent fetal death from cord entanglement, which has been found to occur in up to 97% of monoamniotic twins and in up to 38% of pseudomonoamniotic twins. The notion that cord occlusion is an acute event has led some experts to recommend continuous fetal monitoring of monoamniotic twins in an attempt to prevent fetal death caused by a “cord accident.”
Anecdotally, true continuous monitoring may be difficult to achieve. The success of continuous monitoring can be affected by many factors including: gestational age, patient and fetal movement, maternal obesity, delay in identification of loss of signal, and adjustment of the monitor by labor and delivery staff. A program of continuous fetal monitoring may increase patient anxiety; increases demand on labor and delivery staff; necessitates prolonged bedrest, which may increase the risk of a thrombotic event; and has the potential of increasing medicolegal liability by establishing a standard of care that may not be met.
Given the difficulty of achieving continuous monitoring and the burden it places on patients and staff, the objectives of our study were to determine the percentage of time continuous monitoring of monoamniotic and pseudomonoamniotic twins is achieved for 0, 1, or 2 fetuses and to assess the effect of gestational age on the effectiveness of “continuous” fetal monitoring.
Materials and Methods
This study was approved by the Human Research Protections Program at the University of California, San Diego. Subjects were identified by review of the University of California, San Diego, Multiple Gestation Database. We conducted a retrospective review of all monoamniotic and pseudomonoamniotic twin pregnancies delivered at our institution from 2000 through 2009. The diagnosis of monoamnionicity/pseudomonoamnionicity was made via prenatal ultrasound and was confirmed by placental pathologic evaluation by 1 of 2 trained placental pathologists.
Inclusion criteria were the attainment of at least 24 weeks’ gestation with 2 viable fetuses, confirmation of monoamnionicity or pseudomonoamnionicity by pathologic examination of the placenta, inpatient antenatal monitoring of both fetuses prior to delivery, and delivery at our institution. Exclusion criteria were congenital anomalies that prevented antepartum fetal monitoring or fetal death <24 weeks’ gestation.
To determine the effectiveness of continuous fetal monitoring, the primary outcome measure, 1 of 2 investigators reviewed the entire fetal heart rate tracing for each patient’s antepartum admission using the electronic external fetal monitoring archiving system Centricity Perinatal (General Electric Healthcare, Fairfield, CT). The total number of potential monitoring minutes was calculated for each patient as the time from admission to the time of delivery. Times of successful continuous external fetal monitoring were calculated and grouped into monitoring categories as follows: 0, 1, or 2 fetuses were monitored. A fetus was considered on the monitor after establishing at least 2 minutes of fetal heart tracing, and similarly, was considered off the monitor after 2 minutes of uninterpretable or no fetal monitoring. The duration of successful fetal monitoring in each monitoring category was converted to a percentage of the total antepartum hours.
Additionally, data in each monitoring category were stratified into 3 gestational age groups: <27 weeks’, 27-30 weeks’, and >30 weeks’ gestation. This allowed calculation of the minutes in each gestational age group in which neither, 1, or 2 fetuses were successfully monitored. The gestational age groups were chosen to divide the average antepartum admission approximately into thirds.
The terminal electronic fetal heart rate tracing for the 6 patients delivered for nonreassuring fetal status was categorized into an acute or subacute event. An acute event was defined as a fetal heart rate pattern that prompted delivery within 30 minutes and a subacute event one that prompted delivery >30 minutes from the onset of the abnormal pattern. The fetal heart rate tracings from the antecedent week were reviewed to see if there were harbingers of abnormal fetal heart rate patterns.
Maternal and neonatal medical records were reviewed for potential covariates. These included: age, gravidity, parity, gestational age at and the indication for hospitalization, antenatal length of stay and management, indication for delivery, gestational age at delivery, Apgar scores, neonatal birthweight, neonatal complications, and neonatal length of stay.
Our institutional protocol for management of monoamniotic/pseudomonoamniotic twin pregnancies includes: offering inpatient hospitalization at fetal viability (24 weeks’ gestation) for continuous external fetal monitoring, administration of an antenatal course of betamethasone, and elective cesarean delivery at 34 weeks’ gestation or after determination of fetal lung maturity. Every effort is made to monitor both fetuses as completely as possible: nursing staff assesses the electronic monitors frequently and if the monitor becomes dislodged or removed it is replaced; the nurse to patient ratio is kept at 1:1 or rarely increased to 1:2 for monoamniotic/pseudomonoamniotic twins on continuous fetal monitoring; and patients are kept on bedrest as completely as possible to aid in obtaining continuous fetal monitoring. Intermittent pneumatic compression devices were ordered on all patients for venous thromboembolism prophylaxis.
Point estimates were presented with 2-tailed 95% confidence interval (CI). The continuous data were not normally distributed and as such overall group differences were analyzed with Kruskal-Wallis test with a P value of < .05 to achieve statistical significance. Post hoc comparisons between groups were likewise performed nonparametrically and a Bonferroni correction was applied to control for multiple comparisons, requiring a P value of < .017 to achieve statistical significance. All statistical analyses were performed using software (SPSS v16.0; SPSS, Inc, Chicago, IL).
Results
Seventeen twin gestations (34 fetuses) with a total of 10,402 hours of fetal heart rate tracing were included in the study ( Table 1 ). Thirteen monoamniotic twin gestations were identified. One twin pair included an acardiac fetus that was not monitored and therefore this twin pair was excluded from the study. One triplet pregnancy (patient 8) containing a monoamniotic twin pair was managed as a monoamniotic twin gestation (the third fetus, functionally a singleton triplet, had a large brain mass and was not monitored) and therefore was included in the analysis. There were 3 patients identified (patients 3, 4, and 14) that had a pseudomonoamniotic (functionally monoamniotic) twin gestation and were included in the study. Two were as a result of an incidental septostomy at the time of laser ablation surgery for twin-twin transfusion syndrome and a third had a spontaneous rupture of the intertwin membrane in the midtrimester.
