The etiology of uterine inversion is unknown. Theories can be summarized into predisposing problems, pregnancy contributions, and intrapartum management.

Predisposing problems include congenital anomaly or weakness of the uterine wall (1,2,3), weakness at the site of placental implantation, disturbance in the contractile ability of the myometrium (4), and tumors (1).

Pregnancy contributions include fundal placenta (5), adherent placenta, placenta accreta, short cord, and increased intraabdominal pressure (1,4).

Intrapartum management factors include manual removal of the placenta, or mismanagement of the third stage of labor (including improper cord traction, fundal pressure, and absence of uterotonic agents) (1,4).

As discussed in the following section, these theories have been debated through the literature over time. With a lack of clear etiology for uterine inversion and its spontaneous occurrence in otherwise low-risk women, many have simply emphasized congenital predisposition to uterine inversion (1,2,3), supported by its recurrence in future pregnancies (3).

The incidence of puerperal uterine inversions ranges from around 1 in 2,000 to 1 in 20,000 (6,7,8,9,10). Current data are based on case reports, case series, a case-control study, and two nationwide cohort studies. Both nationwide cohort studies were performed in the last 10 years (7,9).

One of these studies by Coad and colleagues (7) found an incidence of uterine inversion of 2.9 per 10,000 deliveries using
data from the U.S. Nationwide Inpatient Sample from 2004 to 2013. The authors found an increased rate of abnormal placentation in women with uterine inversion (6.3% vs. 0.5%, adjusted odds ratio [aOR] 13.6; 95% confidence interval [CI] 11.5-16.1), prolonged labor (aOR 1.58; 95% CI, 1.12-2.25), and severe preeclampsia (aOR 2.43; 95% CI, 1.98-2.98) (7). They were unable to assess the effect of placental location, labor augmentation, precipitous delivery, uterine anomalies, parity, or the degree of inversion as these variables were not available in the database.

Prior authors have found uterine inversion to be more common in primiparas (2,4,11), with some speculating atony after longer labor as a potential cause of uterine inversion. However, this has subsequently been refuted, and the higher proportion of cases among nulliparous women is thought to be secondary to a higher representation of nulliparous women in the study populations (6). A case-control study found no association between uterine inversion and parity (6).

Mismanagement of the third stage of labor (e.g., excessive traction on the umbilical cord) is often identified as a risk factor for inversion, more heavily in earlier publications, and although difficult to study, there is no clear evidence that this leads to an increased risk of inversion. Das reports that of the acute puerperal cases (217 cases), 40% were of spontaneous origin, 21% from traction on the umbilical cord, and 19% from an improper method of expressing the placenta (4). Although mismanagement of the third stage of labor is difficult to quantify, a retrospective review by Baskett (8) of 40 cases over a 24-year time period split the cases into those that occurred before practice changes to active management of the third stage of labor and those that occurred after (practice changes occurred in the late 1980s) and found that there was a 4.4-fold decrease in the number of inversions after vaginal deliveries (8). Baskett also noted that manual extraction of the placenta (which occurred in 11.1% of their acute puerperal inversion cases after vaginal delivery) was associated with uterine inversion although recognizes that a control group would be required to appropriately investigate this association (8). The need for excessive cord traction or manual removal of the placenta may simply indicate that the placenta is abnormally adherent, resulting in an increased risk of inversion if traction is applied. We recommend consideration of the possibility of an abnormally adherent placenta, should the placenta not deliver easily. However, active management remains appropriate to prevent other complications including postpartum hemorrhage.

Nonpuerperal inversion is even less common than puerperal inversion. The majority of nonpuerperal inversions are benign myomas, with a minority from malignant tumors, most commonly a leiomyosarcoma, and normally occur in older women (12). Although this chapter does not delve into nonpuerperal inversion, it is important to be aware of its existence because some surgical techniques we discuss were documented in nonpuerperal cases. Similar to subacute and chronic puerperal cases, the cervical contraction of nonpuerperal cases limits nonsurgical management.

Uterine inversion should always be considered in the differential diagnosis when called to the bedside for postpartum hemorrhage, hypovolemic shock, or abdominal pain. In Das’ review of 391 cases (n = 297 puerperal), hemorrhage or shock was present in all but four cases (4). In Johnson’s review of nine cases, all but one had hemorrhage and shock (13). In Watson’s retrospective cohort review with 18 cases of inversion, hemorrhage occurred in 94% of cases and shock in 39% of cases (6). In all 18 cases, the placenta was separated from the uterine wall before replacement. Not surprisingly, blood loss correlated with the duration of inversion and shock. Traction of the peritoneum, adnexa, and neurologic plexus in the broad ligament are also thought to contribute to shock (2,4).

On physical examination, the fundus will not be palpable abdominally, a vaginal mass will be palpated or visible. If uterine inversion occurs in the third stage of labor during a vaginal or cesarean delivery, the placenta will often remain attached to the inverted uterus. An acute presentation is easier to identify than a subacute or chronic inversion. The provider may also feel a contracted cervical ring surrounding the vaginal mass, note that the vaginal mass is friable and bleeding, feel a cup-like indentation instead of a palpable fundus, or fail to feel the uterus in its anatomical position on a rectal examination. For the less obvious cases of subacute, chronic, or first-degree inversions, imaging (which is discussed below) can aid in diagnosis. In the majority of cases, however, one should rely on clinical signs and symptoms for diagnosis in the setting of hemorrhage, shock, or abdominal pain combined with the above physical examination findings.

Similar to postpartum hemorrhage, the differential diagnosis in the acute time period includes uterine atony, retained placenta, lacerations, coagulopathy, or other vaginal or uterine mass.

In the subacute or chronic time period, the differential includes a retained placenta, uterine fibroid, endometrial or cervical polyp, vaginal hematoma, vaginal cyst, or solid mass.

Initial management should be focused on maintaining hemodynamic stability. The provider should immediately alert others to the complication of uterine inversion and get additional staff members and anesthesiology support. Obtain a second large-bore IV (14 or 16 gauge size), make sure a type and screen are available, and request a type and cross for blood products if profusely bleeding. Immediately give a 1-L fluid bolus if in shock with additional boluses as necessary while awaiting replacement blood products.

After evaluation to confirm the diagnosis, or if recognized acutely during the third stage, the most critical step is a prompt replacement of the uterus. To facilitate replacement, stop uterotonics, explain the findings to the patient, and ensure anesthesia is at bedside for support of pain control and possible administration of uterine relaxation agents.

Although there may be controversy on replacement of the uterus compared to placental removal first followed by replacement (1,4) in the acute period, we recommend an immediate attempt at replacement first. If manual replacement fails owing to the size of the placenta, removal can be considered before ongoing attempts. However, this should be done with caution as the attached placenta may prevent bleeding from the placental bed. Immediate attempt at uterine replacement cannot be emphasized enough as this is critical to reducing blood loss and preventing shock (6,11). After uterine replacement and uterotonics, attention can then be turned to exploring the uterine cavity to remove the placenta in a controlled environment.

A specific method for manual replacement was described in nine cases of successful replacement by Johnson in 1949. His described method is based on the concept that replacement is dependent on the uterine ligaments and that it is necessary to replace that which more recently inverted first. He describes his method as follows: Place your hand in the vagina with the fundus in your palm and the fingertips on the uterocervical junction, apply pressure with your fingers to first widen the cervical ring, then push upward to lift the uterus above the level of the umbilicus to return the ligaments to their anatomical position; lifting the uterus out of the pelvis allows the passive action of the broad ligaments to assist in
correction with the fundus returning last to the abdominal cavity (Figure 5.3.1). He describes feeling the fundus receding from the palm of the provider’s hand. After the return of the uterus to the abdominal cavity, hold the position for 3 to 5 minutes with the entire hand and most of the forearm in the vagina to ensure sufficient ligament tension for correction (13). In his series, Johnson used general anesthesia in all cases, with successful reversion in all of them and two going on to have future pregnancies.

In 1948, Henderson and Alles described using ringed forceps on the cervical ring for countertraction during manual replacement of the fundus (2). In their retrospective review of 24 cases, they had 6 deaths which they believe was related to a delay in diagnosis and insufficient resuscitation (2). Although they do not report how many of their cases they attempted to replace manually first (without ringed forceps), they report that all their cases successfully replaced the uterus without additional surgical management (2).

O’Sullivan described two cases in 1945 in which hydraulic pressure was used to replace the uterus. In his publication, he used a douche preparation held 2 to 3 feet above the vagina, most likely to increase the pressure of the preparation while blocking the vaginal outlet with both forearm and hands; this distended the vagina and pushed the uterus back to its correct position (Figure 5.3.2) (14). This method was successfully used in a case report published by Ward in 1998 using normal saline for hydrosufflation (15).

Uterine relaxation may be needed to replace the uterus. Nitroglycerin (a smooth muscle relaxer) may be required for uterine replacement. Around 100 µg can be given by the anesthesiologist without causing hypotension (16). Ongoing fluid resuscitation should be provided by the anesthesiologist to maintain hemodynamic stability.

After replacement of the uterus, perform manual exploration to rule out a predisposing or iatrogenic uterine rupture and remove the placenta if not already removed. A Foley catheter should be placed in the bladder to avoid urinary retention. Uterotonics should be given to assist in ensuring appropriate uterine contraction.

If the above nonsurgical methods fail, cervical contraction is likely the limiting factor, and a surgical approach is necessary.