Uterine inversion is a rare complication of the third stage of labor but is potentially life-threatening. Although largely preventable, some occurrences are unavoidable. Prompt recognition and management are critical to reduce maternal morbidity and mortality rates, mainly due to hemorrhage.

The classification systems of uterine inversion are based on either the duration or magnitude of the inversion. Criteria are found in Table 31-1, and examples are seen in Figures 31-1 and 31-2 (Kitchin, 1975; Livingston, 2007; Pauleta, 2010; Watson, 1980; You, 2006). Most are acute and second- or third-degree inversions (Baskett, 2002; Brar, 1989; Dali, 1997; Morini, 1994; Platt, 1981; Shah-Hosseini, 1989; Witteveen, 2013).

The reported incidence of uterine inversion varies widely, which may be due to differences in definition, patient populations, and awareness and recognition. The reported incidence ranges from 1 in 500 to 1 in more than 57,000 deliveries (Baskett, 2002; Bunke, 1965; Das, 1940; Hostetler, 2000; Morini, 1994; Shah-Hosseini, 1989; Watson, 1980; Witteveen, 2013). Two single-institution reports that analyzed long epochs cite incidences of 1 in 1860 during cesarean delivery, 1 in 3737 during vaginal delivery, and 1 in 6403 in all delivery settings (Baskett, 2002; Shah-Hosseini, 1989). In a nationwide population-based study, the incidence was 1 in 20,000 vaginal births (Witteveen, 2013).

Previously, uterine inversion during cesarean delivery was considered rare (Chatzistamatiou, 2014; Witteveen, 2013). However, in one series, the incidence of inversion during cesarean delivery was actually twice that associated with vaginal delivery (Baskett, 2002). As one explanation, management of third-stage labor during cesarean delivery has varied over time and ranged from immediate manual extraction of the placenta to cord traction to promote spontaneous expulsion, which is an inversion risk. Second, inversion during cesarean delivery is typically recognized and managed more expediently than at vaginal delivery. Accordingly, rates of significant clinical complications such as severe hemorrhage and blood transfusion are lower (Baskett, 2002; Terp, 1998).

Several items have historically been associated with uterine inversion, although no particular factor has been demonstrated conclusively to pose significant risk. Mismanagement of third-stage labor may include excessive cord traction and the Credé maneuver (Das, 1940; Kitchin, 1975; Watson, 1980; You, 2006). With this maneuver, the abdomen is grasped by one hand so that the thumb rests on the anterior fundus and the fingers on the posterior fundus, while firm, steady pressure is directed caudally in the axis of the superior strait. Other reported factors are a short umbilical cord, fundal placental implantation, congenital weakness of the uterine wall, rapid labor and delivery, oxytocin or magnesium sulfate use, and primiparity (Adesiyun, 2007; Brar, 1989; Das, 1940; Kitchin, 1975; Platt, 1981; Witteveen, 2013). Importantly, although mismanagement of third-stage labor is often implicated, this association remains unproven (Shah-Hosseini, 1989; Watson, 1980). In fact, active management of this labor stage may be protective. Baskett (2002) described an active management of third-stage labor that consisted of administration of oxytocin (5 units intravenously [IV] or 10 units intramuscularly [IM]) after delivery of the anterior shoulder. Implementation of this protocol reduced the incidence of uterine inversion fourfold compared with a prior epoch in which active management was not used.

Whether retained placenta or placenta accreta, or its variants, are associated with uterine inversion is unclear. Specifically, morbidly adherent placenta, which is another term for placenta accreta, is described as a risk for inversion in case reports and in older series. However, this link has not been specifically addressed in more recent articles (Agarwal, 2005; Das, 1940; Hostetler, 2000; Kitchin, 1975). For example, manual removal of the placenta is a suggested risk from newer studies, but a specific association with this maneuver for an undiagnosed morbidly adherent placentation has not been analyzed (Baskett, 2002; Kitchin, 1975; Shah-Hosseini, 1989; Watson, 1980). Moreover, although the prevalence of placenta accreta has risen, a specific association with uterine inversion is described only in case reports (Agarwal, 2005; Tsivos, 2009). Yet, despite a lack of extensive data regarding a link, placenta accreta, particularly if unrecognized, may predispose to uterine inversion and should be kept in mind. This is particularly important in light of the higher cesarean delivery rate in current obstetric practice.

The signs and symptoms of uterine inversion depend to some degree on its acuity and extent. Classically, bleeding, pain, and an intravaginal or protruding mass are found, and the fundus cannot be palpated abdominally. Since most cases are acute and complete, severe hemorrhage is typical and leads to shock with hypotension, altered mental status, and cool, pale skin. Historically, “shock out of proportion to blood loss” has been described. This degree of shock has been attributed to stretching of the pelvic parasympathetic nerves and subsequent increased vagal stimulation. However, no solid evidence supports this supposition. Instead, the degree of shock more likely represents an underestimated blood loss (Baskett, 2002; Beringer, 2004; Das, 1940; Hostetler, 2000; Kitchin, 1975; Platt, 1981; Shah-Hosseini, 1989; Watson, 1980).

Incomplete uterine inversion, although less common, may be more challenging to detect (Morini, 1994). Bleeding may be lighter. Also, vaginal examination may be less clear due to absence of a large vaginal mass and the possibility of still being able to palpate part of the fundus abdominally. For these reasons, incomplete inversion is more likely to be associated with the subsequent diagnosis of subacute or chronic inversion. With chronic inversion, urinary retention may be another finding (Thakur, 2014).

The vast majority of uterine inversion cases are diagnosed clinically, and life-saving management should not be delayed to obtain confirmatory imaging. Imaging may be most helpful with incomplete inversion or with a delayed diagnosis. Sonographic findings for both incomplete and complete inversion include a “mirror sign,” in which the uterus has a U-shaped cavity due to the inverted fundus extending to the cervical os. A “pseudostripe” can be formed by the opposing uterine serosal surfaces. Also, a “target sign” represents the hyperechoic inverted fundus, which is surrounded by the hypoechoic lower uterine segment (Pan, 2015; Rana, 2009; Thakur, 2014). Another sign is the appearance of the ovary(ies) at the indentation site of the inverted uterus (Smulian, 2013). Three-dimensional sonographic imaging has also been used to diagnose incomplete inversion (Pauleta, 2010). In uncertain cases, magnetic resonance (MR) imaging can provide additional anatomic detail (Thakur, 2014). T1-weighted images are usually uninformative, but T2-weighted images can show an inverted cavity. The adnexa and round ligaments may appear to be drawn into this cavity. In addition to diagnosis, sonography can confirm uterine repositioning (Smulian, 2013).

General Considerations

Effective management of uterine inversion requires prompt recognition and quick mobilization of resources to minimize maternal morbidity and mortality. If immediate manual replacement fails, transfer to the operating room is considered early in the process to allow hemodynamic monitoring, to restore intravascular volume by administration of fluids and blood products, and to manage hemorrhage. Anesthesiology staff is ideally immediately available. They can assist with volume resuscitation and administer uterine relaxants or tocolytics. They may be called upon to provide general anesthesia for additional manual vaginal manipulations or for rapid conversion to laparotomy.

As outlined in Figure 31-3, general principles of management include:

1. 
   

   Promptly recognize and notify support personnel, operating room staff, and anesthesiology providers

   
   
2. 
   

   Withhold uterotonic agents and immediately attempt manual replacement of the uterus with the placenta attached, if possible

   
   
3. 
   

   Administer uterotonic agents if the uterus is successfully repositioned to prevent further atony and repeat inversion

   
   
4. 
   

   Respond early and aggressively to ongoing hemorrhage. This will include serial assessment of vital signs, placement of large-bore IV lines, infusion of crystalloid solution in a 3:1 ratio to estimated blood loss, and activation of postpartum hemorrhage transfusion protocols

   
   
5. 
   

   Obtain blood samples to monitor ongoing blood loss and identify potential disseminated intravascular coagulopathy. Tests include complete blood count (CBC), prothrombin time (PT), partial thromboplastin time (PTT), international normalized ratio (INR), and haptoglobin and fibrinogen levels.

Recent data from the battlefield and trauma centers have shown decreased mortality rates in nonobstetric populations when massive transfusion protocols are used. These involve fresh-frozen plasma:packed red blood cell (FFP:pRBC) ratios of 1:1, 1:1.5, or 1:1.8 (Gonzalez, 2007; Riskin, 2009). Components to be considered and massive transfusion protocols are addressed in greater detail in Chapter 7 (p. 98).

Nonsurgical Management

Manual Repositioning

As noted, uterine inversion may occur during vaginal or cesarean delivery, and general steps to reduce it are similar. First, replacement of uterine inversion should be attempted immediately after diagnosis. Manual replacement of the uterus, termed the Johnson maneuver, involves pushing the inverted fundus toward the umbilicus.

With an inversion after vaginal delivery, the fundus is elevated through the contracted myometrial ring, which forms at the upper cervix, to restore normal positioning (Figs. 31-4 and 31-5) (Johnson, 1949). Placing ring forceps on the cervical ring may aid effective countertraction (Henderson, 1948; Kitchin, 1975). If the placenta is still attached, it should remain in place until after reduction of uterine inversion. Premature placental removal can worsen bleeding from the placental site (Watson, 1980).

If an initial manual replacement attempt is unsuccessful, administration of a tocolytic agent can induce myometrial relaxation. This can assist further uterine manipulation and potentially avoid more extensive surgery. Suitable options are terbutaline, ritodrine, nitroglycerine, or magnesium sulfate. The last is preferred if hypotension is present. Dosages are listed in Table 31-2. If still unsuccessful, further attempts at manual reduction are made with the patient under general anesthesia. Abouleish and coworkers (1995) reviewed anesthetic management in 18 cases of acute uterine inversion. Manual repositioning of the uterus was successful in 4 cases (22 percent) without need for any tocolytic agent. Eight women received a single dose of 0.25 mg IV terbutaline, with 5 (63 percent) achieving successful uterine reduction. General anesthesia was required in the other 3 cases. The remaining 6 patients underwent general anesthesia as the initial management approach.