Intrauterine hemostatic balloon placement: is <12 hours really better?

We commend Einerson et al for answering a very important unsolved question: whether the Bakri balloon can be removed <12 hours after placement. Their retrospective observational study showed that Bakri balloon removal before or after 12 hours does not affect outcomes, and thus concluded “if ongoing hemorrhage has abated, it is reasonable to consider removal of an intrauterine balloon by 12 hours after its initial placement.” We have a concern and an addition.

Our concern regards selection bias. Who decided early (<12 hours) vs late (>12 hours) removal? They stated, “the duration of balloon was at the discretion of the clinical provider and was not dictated by protocol.” They also stated, “the study was performed at a single academic institution with a high obstetric volume, a postpartum hemorrhage protocol, and experience with balloon.” It is therefore likely that experienced obstetricians (“experts”) decided. Although the background characteristics did not differ between the early and late groups (their Table 1), this does not eliminate selection bias. For example, uterine contraction usually affects decision-making by an expert. If uterine contraction becomes better after balloon placement, there is no need for prolonged placement, but if the uterus contracts but occasionally becomes floppy (repeatedly floppy contraction), it may be better to place it longer. There may be no difference demonstrable in Table 1 to differentiate these 2 situations. In addition to uterine contraction, experienced obstetricians evaluate early vs late removal by intuition, which may involve, for example, bleeding pattern, placental location, and placental separation pattern. The basis of this intuition should be analyzed and determined, so that less experienced obstetricians can utilize this experience. The data of Einerson et al should be interpreted as, “it is reasonable to consider removal of an intrauterine balloon by 12 hours based on the judgment of an experienced obstetrician to the extent that they can judge the merits of early vs late removal in this patient.”

Our addition regards balloon prolapse. While the balloon remained intrauterine in 274 patients (study population), it was prolapsed in 33. Balloon prolapse should be prevented, and is preventable. We devised “holding the cervix” (closing the cervical ostium with forceps, preventing balloon prolapse), “abdominal traction stitch” (balloon shaft being pulled cephalad through the abdominal wall), and their combination to achieve this goal. Depending on the situation, we use either procedure, preventing balloon prolapse. The outcome of these 33 patients with balloon prolapse is of interest.