Mobile Phone Application Versus Handheld Device for Fetal Vibroacoustic Stimulation

Overview

This is a randomized controlled single blinded crossover study to compare fetal vibroacoustic stimulation with a handheld device to a mobile phone application designed for this study. The handheld device, the Corometrics model 146, was previously marketed and sold for fetal stimulation and is modeled after devices used as artificial larynxes. The mobile phone application uses an iPhone's built in vibration with acoustic stimulation from the speaker at 2,000 Hz frequency at a sound level of 74 decibels when measured at 1 meter distance. The investigators' hypothesis is that the mobile phone application will elicit a fetal response (defined as one or more fetal heart rate accelerations in 15 minutes) at a similar rate to that of the handheld device. Fetal vibroacoustic stimulation is commonly used during antenatal tests of fetal well being such as a non-stress test. Fetal vibroacoustic stimulation works by arousing the fetus to a state of wakefulness during which reassuring fetal movements and associated fetal heart rate accelerations occur. During a non-stress test, two or more fetal heart rate accelerations in 20 minutes constitutes a reassuring test. Fetal vibroacoustic stimulation has been shown to decrease the false positive rate of non-stress testing without increasing the false negative rate. This study will compare the frequency that one or more fetal heart rate accelerations occur in the 15 minutes after vibroacoustic stimulation with the Corometrics-146 fetal acoustic stimulator compared to after vibroacoustic stimulation with a mobile phone application designed for the study.

There are multiple proposed methods of assessing fetal well-being in utero. One such method which has been well accepted for fetal surveillance is the non-stress test (NST), during which the fetal heart rate is monitored with cardiotocography for a period of 20-40 minutes, depending on the time it takes to confirm reactivity. Reactivity is confirmed when there is an increase in fetal heart rate, called an acceleration, the presence of which has been shown to reliably predict the absence of fetal metabolic acidemia. Non-stress testing assumes that with fetal movements, a nonacidotic, neurologically normal fetus will have an acceleration. Fetuses develop reactive tracings between 28 and 32 weeks, with 80% of normally developed fetuses displaying reactive fetal heart rate tracings by 32 weeks. Two limitations to fetal non-stress testing have been recognized. Due to fetal sleeping periods during which movement is decreased, tests may be falsely interpreted as non-reactive, which can lead to unnecessary intervention such as prolonged monitoring, additional testing such as the biophysical profile, induction of labor, or delivery by cesarean section. In addition, if a tracing is not found to be reactive after 20 minutes, extending the testing time to 40 minutes may eventually produce a reactive tracing. This, however, is taxing on antepartum unit equipment, staff resources, and can create additional patient anxiety. Vibroacoustic stimulation (VAS) is a method used to arouse the fetus from a state of fetal sleep to fetal activity, reflected as an increased heart rate, called an acceleration, on electronic fetal monitoring. Until now, VAS has been emitted from an artificial larynx called a vibroacoustic stimulator which is held on or above the pregnant woman's abdomen to transmit audible frequency to the fetus. For example, one such device, the Corometrics-146, sends out a vibratory sound stimulus with a frequency of approximately 75 Hz and an intensity of approximately 74db for a duration of 3 seconds. A recent Cochrane Review concluded that the use of VAS in antenatal testing reduced both incidence of non-reactive fetal heart rate tracing and overall testing time. This has implications for clinical use, allowing equipment and staff resources to be available to test more women with less inconvenience for the patient. Many antenatal testing units only have one VAS device, and these can cost up to $500. In addition, the Cochrane Review highlighted a need for further research into the realm of vibroacoustic stimulation in obstetric care, with particular attention to optimum intensity, frequency, duration, and position of the stimulation. Current devices, however, only come with one set frequency and intensity, making studies outlining optimal response rates with varying acoustic parameters impossible. The investigators believe that the development of a mobile device (such as a phone) application which produces a vibroacoustic sound wave similar to that produced by the artificial larynx can address the aforementioned issues. This would make vibroacoustic stimulation more available, and allow for further testing with differing parameters such as frequencies, intensities, and duration of stimulation, which are easily adjusted with an application, but not an artificial larynx. Patients who are already in the hospital for antepartum fetal monitoring will be recruited and consented. The first 10 patients meeting inclusion criteria will be involved for the purposes of testing for interference between the vibroacoustic stimulation devices and the ultrasound transducer which is monitoring the fetal heart tracing. This process will consist of recording a fetal heart rate tracing while stimulating the fetus with the VAS device 12 inches from the transducer without touching the maternal abdomen, 12 inches from the transducer while in contact with the maternal abdomen, and right next to the transducer. Any consistent change in tracing at the time of stimulation will be reviewed. Once it is determined that there is no interference, the data collection will continue. Data will be collected from pregnant patients who present to triage, antepartum floor, and labor and delivery at 32 weeks or greater gestation. Once a patient's fetal heart rate tracing is not reactive for 5 minutes they will be randomly assigned method of vibroacoustic stimulation either by the Corometrics-146 artificial larynx or mobile phone application. A recording for heart rate accelerations will be taken for 15 minutes and scored later by a blinded observer. Patients will then undergo stimulation with the other device followed by recording of heart rate tracing for a further 15 minutes, for a total of 35 minutes on the monitor. Patients who present to the hospital on more than one occasion will be able to participate up to 5 times if they express continued interest in participating in the study. The primary outcome will be the percentage of the time that one or more fetal heart rate accelerations occur within 15 minutes after vibroacoustic stimulation. In this crossover study each patient will receive vibroacoustic stimulation with each of the two devices with the order determined at random. The mother will be partially blind to the method of stimulation by having each device in an opaque bag. Complete blinding of the mother will not be possible since the two devices sound and vibrate differently. Fetal heart rate tracings will then be reviewed by an obstetrician with expertise in cardiotocography interpretation who are blinded to the method of vibroacoustic stimulation for each tracing. An interim analysis will be performed once 100 patients (50 in each arm) have been enrolled. At this time the power analysis will be re-calculated to confirm the number of patients required to meet statistical significance. An analysis will be done to ensure that the state of the fetus has returned to the baseline state before vibroacoustic stimulation with the second device. This temporary change in fetal state of wakefulness following vibroacoustic stimulation and return to a baseline fetal state of rest has been described in the literature; the effect of the vibroacoustic stimulation seems to wash out over a period of time. The blinded expert will determine for each patient if the effect of vibroacoustic stimulation with the first device has washed out before stimulation with the second device. This will be done by determining if the 5 minute tracing before stimulation with the first device appears similar to the 5 minute tracing before stimulation with the second device. If the effect of the first device is observed to wash out in 75% or greater of the patients the study will continue as a cross over study. If not the study analysis will be done with each patient receiving stimulation with only one device chosen at random (and analysis of the first 100 patients only using data from the first device). The investigators determined that the sample size will need to be at most 400 patients, 200 per group, to account for any dropouts, with a goal of obtaining data on 175 in each group, for a total data set of 350 patients. This would provide a power of 80% to detect a 15% difference in the rate of the primary outcome from a baseline rate in the range of 40-80%, at a two-sided alpha level of 0.05 using chi-squared test. If a wash out effect is observed in greater than 75% of the patients the study will continue as a crossover study and fewer patients will be needed.