In this pilot study, use of the electromechanical IPD device significantly reduced the risk of full LAM avulsion—no participant in the Device-treatment group experienced full LAM avulsion. This result supports the concept that the elasticity of the pelvic floor muscles can be gradually enhanced through mechanical stretching prior to delivery of the fetal vertex, resulting in reduced likelihood of LAM avulsion from its attachment at the pubic rami. This is consistent with the findings of a previous feasibility study using an earlier version of the device [12].

The mechanism for reducing pelvic floor injury with use of the IPD device is based on the properties of the pelvic floor muscles [20, 21]. Research in sports medicine indicates that muscle and connective tissues are viscoelastic and elasticity is strain-rate dependent, i.e., the slower a muscle is stretched, the less likely it is that there will be injury [22]. In the second stage of labor, the pelvic floor muscles experience rapid distension. MRI-based computer modeling shows that the distal-most pelvic muscles comprising the levator complex must lengthen by a factor of 3‒4 during fetal head crowning [10, 11]. Classic teaching in obstetrical care emphasizes the role of the clinician in limiting the rate of pelvic floor stretch during delivery by supporting the perineum and controlling the speed of head expulsion [23]. Paradoxically, the extreme dilation of the pelvic tissues during childbirth creates resistance that impedes and slows delivery of the baby, and simultaneously over-stretches and injures pelvic tissues. The currently investigated electromechanical IPD device slowly prepares the pelvic floor muscles by maximizing their stretch prior to delivering the fetal head.

A major benefit of reducing levator ani injury risk during vaginal delivery is the probable reduced risk of developing POP later in life [3, 6, 7], an association that appears to strengthen over decades [24]. Indeed, 55% of women with LAM avulsion develop POP within 6‒17 years after first delivery compared with 21% of women without LAM avulsion [3]. Injury to the LAM may also increase the likelihood of developing stress incontinence, although that relationship is equivocal [25]. To exclude potential pre-existing pelvic floor injury in our cohort, we only enrolled nulliparous individuals. Second vaginal deliveries do not seem to have a deleterious effect on LAM biometry and structural integrity [2, 26]. Future longitudinal studies will determine if the IPD device can be safely and effectively used in second pregnancies, or whether first pregnancy IPD treatment provides adequate protection against LAM avulsion injuries in subsequent vaginal deliveries.

The IPD device had a favorable safety profile when used as an adjunct during the first vaginal delivery. Nonserious vaginal abrasion/bruising identified in five Device participants (6.1%) was considered a device-related minor AE; there were no cases of vaginal abrasion or bruising reported in the Control arm. The study protocol instructed investigators to examine the vaginal walls for any evidence of tissue injury after device removal. This may have introduced bias because no intrapartum examination of the vagina was required for the Control arm.

A significant difference was recorded between the two groups for SAE occurrence. However, of the 16 participants in the Device group with a maternal SAE, 3 were preeclampsia with severe features, 2 were cases of retained placenta, and 3 were related to non-obstetrical infection. None of these complications was device related. There was no significant difference in the incidence of L&D- and device-related SAEs between the Device and Control groups (14.6% vs 6.9% respectively, p = 0.09).

The only device-related SAE reported was a vaginal laceration seen after device removal that required extensive suturing to achieve hemostasis. Of note, that participant had a nonreassuring fetal heart tracing that required maternal lateral repositioning multiple times. In addition, her vaginal mucosa was very friable and denuded, which contributed to the challenges in controlling bleeding. Subsequently, the instructions for use of the device were changed to limit the number of maternal repositioning maneuvers with the device in place and to exclude individuals identified with friable vaginal tissue prior to randomization.

The rates of chorioamnionitis and endometritis were statistically similar in the Device and Control arms, suggesting that the presence of an intravaginal device in labor does not enhance the risk of ascending bacterial microbes and infection. The number of cervical examinations performed during labor is an independent risk factor for the development of clinical chorioamnionitis [27]. Placement of the IPD does not require a digital examination of the vagina or cervix. Once the device is placed, there is no requirement for additional internal examinations. Unlike intrauterine pressure catheters that are inserted through the cervical canal and into the uterine cavity providing a potential pathway for ascending bacteria [28], the IPD is placed only 4‒5 cm within the vaginal canal.

We theorized that pre-stretching of the pelvic floor with the device would mimic the perineal massage technique that has been shown to potentially reduce the risk of serious perineal lacerations [29]. The incidence of ultrasound-diagnosed OASIS in the Device and Control arms was similar. Taken together with the significant reduction in complete LAM avulsion rate, the IPD device appears effective for stretching and preparing the pelvic floor muscles to accommodate rapid delivery of the newborn, but does not have a discernable impact on perineal/anal sphincter injury rates. Participants who employed the IPD device reported favorable impressions and experiences, with no negative impact on quality of life.

A primary study limitation was the mid-study protocol change to a more stringent primary efficacy outcome (i.e., full avulsion rate instead of combined partial + full avulsion rate). Although this change invalidated initial power calculations of sample size requirements, this pilot study was nonetheless able to discern a statistically significant benefit of treatment on reducing the incidence of full avulsions. The reasoning behind this change was that low-grade partial avulsions may potentially heal postoperatively and thus have limited clinical significance, whereas a fully avulsed LAM is unlikely to spontaneously reattach after injury. A pivotal clinical trial (Clinicaltrials.gov identifier NCT03973281) is currently enrolling subjects to confirm the findings of this pilot study. Another limitation was the necessary partial-blinding design, which may have introduced unintentional bias in recording peripartum events, including AE rates. This did not impact the blinded assessment of LAM avulsion on ultrasound images. Several Device participants converted to C-section or experienced rapid progression to full cervical dilation before randomization; thus, LAM injury findings of this study are not generalizable to the subgroup of individuals with a rapid first-stage labor.

Study recruitment was aimed at including populations with a history of receiving inequitable care for pelvic floor disorders [13, 30]. Participants approximated the proportion of Black individuals and exceeded that of Asian people currently present in the diversifying USA populace [13], and our findings of reduced LAM avulsion with IPD use are likely to apply to a diverse group of labor patients. A study currently ongoing with larger planned enrollment should permit detailed subgroup analyses.

In conclusion, the electromechanical IPD device reduced the prevalence of full LAM avulsion, a surrogate endpoint for developing POP. Counseling pregnant individuals on the risk of long-term sequelae of vaginal delivery is not typically provided during antenatal care. Other than avoiding use of obstetrical forceps, there are few tools available for preventing LAM avulsion. Incorporating the IPD into obstetrical practice may prove beneficial for protecting against pelvic floor injuries responsible for future POP.