Objective
We sought to determine whether the timing of balloon tamponade (BT) and uterine artery embolization is associated with morbidity among women with postpartum hemorrhage (PPH).
Study Design
This is a cohort study of women with PPH at a single tertiary academic institution. Patient demographics, delivery outcomes (eg, labor augmentation, route of delivery), and PPH-specific data (eg, estimated blood loss at the time when PPH interventions were employed) were abstracted from the chart. Outcomes studied included nadir hemoglobin, packed red blood cell transfusion, intensive care unit admission, and hysterectomy.
Results
During the study period, 420 women were eligible for analysis. Women receiving BT at lower estimated blood loss quartiles had higher nadir hemoglobin, less frequent packed red blood cell transfusion, fewer intensive care unit admissions, and fewer hysterectomies. There were no associations found between timing of uterine artery embolization and maternal outcomes.
Conclusion
Earlier use of BT among women experiencing a PPH is associated with decreased maternal morbidity.
Reducing maternal severe morbidity and mortality worldwide is a public health priority. With regard to maternal mortality, there have been considerable advances worldwide over the past 15 years, as evidenced by a 50% decline from 1990 through 2010. This decline, however, has not been seen in the United States. Moreover, one major contributor to maternal mortality and severe morbidity is postpartum hemorrhage (PPH) and the frequency of this adverse event appears to be increasing. For example, Berg et al compared 2 time periods (1993 through 1997 and 2001 through 2005), and found that the rate of PPH increased significantly. Thus, PPH is a key focus in the effort to decrease maternal morbidity and maternal mortality.
The causes of PPH are myriad and management is often multimodal. Common interventions include fundal massage, uterotonic medications, balloon tamponade (BT), uterine compression sutures, uterine artery embolization (UAE), and dilation and curettage. Although studies have shown a decrease in maternal morbidity from implementation of PPH protocols, few studies have been able to demonstrate whether earlier initiation of specific interventions improves outcomes or whether an optimal time to implement interventions can be identified. One study by Zhao et al found that patients who received interventions for PPH at an estimated blood loss (EBL) of 500-700 mL, when compared to those with an EBL of >700 mL, had a lower mean EBL, a lower frequency of blood transfusion, and a higher mean fibrinogen. However, this study only included women who underwent cesarean and received at least 2 interventions starting at an EBL lower than that meeting the standard criteria for PPH.
Consequently, it remains uncertain whether and to what degree the earlier use of intervention in the setting of PPH improves outcomes. The present study was designed to determine whether the timing of BT and UAE is associated with morbidity among women with PPH.
Materials and Methods
This study was approved by the Northwestern University Institutional Review Board. Eligible patients were all those who had a PPH during 2011. These women were identified by querying the electronic medical record to identify those who had an EBL of >500 mL at vaginal delivery, >1000-mL EBL at cesarean delivery, or receipt of blood products or uterotonic medications. The charts of women who had been identified were then reviewed to confirm the diagnosis of PPH and perform data abstraction. During the time of the study, a PPH protocol was in place at the institution such that in the event a PPH was diagnosed and clinically significant bleeding continued, a flow sheet identifying the time at which different EBL levels were reached and when specific interventions, such as BT and UAE, were instituted was routinely utilized. The BT device universally used was Bakri (Cook Medical, Bloomington, IN).
Patient demographic characteristics, delivery outcomes (eg, mode of delivery, receipt of blood products, intensive care unit [ICU] admission, hysterectomy), and PPH-specific data (eg, EBL at different times after the PPH was diagnosed, the EBL at which each hemorrhage intervention was employed) were abstracted from the patient chart. The timing of each intervention, as a function of the EBL at which it occurred, was divided into quartiles. Groups were compared according to these quartiles with regard to the outcomes experienced.
Outcomes that were examined included nadir hemoglobin, packed red blood cell transfusion, ICU admission, hysterectomy, and maternal death. Statistical comparisons were performed with the nonparametric test for trend and Fisher exact test. P < .05 was used to define statistical significance and all tests were 2-tailed. Analyses were conducted using software (STATA, version 11; Stata Inc, College Station, TX).
Results
During the study period, there were 12,166 deliveries, of which there were 797 PPH (6.5%), with 420 PPH flow sheets completed. Characteristics of the 420 women are presented in Table 1 . The majority of the sample was non-Hispanic white, nulliparous, and presented in spontaneous labor. A minority had chorioamnionitis or had been treated with magnesium sulfate. Approximately half delivered vaginally. The mean EBL for the women analyzed in this study was 1477 mL (SD 1178 mL). The mean EBL during cesarean deliveries and spontaneous vaginal deliveries was 1879 mL (SD 1261 mL) and 1045 mL (SD 753 mL), respectively. The etiology of PPH was most commonly uterine atony.
| Characteristic | Value |
|---|---|
| Age, y | 32.0 ± 5.7 |
| Race/ethnicity | |
| Non-Hispanic white | 57.3 |
| Black | 8.6 |
| Hispanic | 20.5 |
| Asian | 6.7 |
| Other | 6.9 |
| BMI, kg/m 2 | 31.2 ± 7.1 |
| Nulliparous | 59.3 |
| Gestational age, wk | 38.7 ± 2.7 |
| Induction | 28.6 |
| Labor augmentation | 41.4 |
| Chorioamnionitis | 13.1 |
| Labor length, h | 13.0 ± 7.7 |
| Magnesium sulfate | 10.5 |
| Birthweight, kg | 3.4 ± 0.7 |
| Route of delivery | |
| Spontaneous vaginal | 43.9 |
| Operative vaginal | 9.3 |
| Cesarean | 46.8 |
| Primary etiology of PPH | |
| Atony | 76.7 |
| Accreta | 1.7 |
| Vaginal laceration | 6.2 |
| Retained products | 6.0 |
| Surgical laceration | 5.0 |
| Uterine inversion | 0.2 |
| Not specified | 4.2 |
Of the 420 women, 11.4% and 4.8% underwent BT and UAE, respectively. The EBL at which each intervention was used, defined by quartile, as well as the associations between the time of intervention and adverse outcomes, are presented in Table 2 . There was evidence for BT that several outcomes were improved when the intervention was initiated at an earlier EBL. For example, women who received BT at an earlier EBL had higher mean nadir hemoglobin, and were less likely to receive red blood cell transfusion, be admitted to the ICU, or undergo a hysterectomy. Conversely, timing of UAE did not seem to make a difference with regard to outcome analyzed. Given the infrequency of UAE, timing of initiation was additionally defined by its occurrence above or below the median EBL. When the chance of hysterectomy for women undergoing UAE was compared, not by quartile, but according to whether or not the UAE occurred below or above the median EBL (ie, the bottom 2 quartiles compared to the upper 2 quartiles), women with UAE at a lower EBL had a decreased risk of hysterectomy (0% vs 38%; P = .021). There were no maternal deaths in the cohort.
