Cost analysis of intrauterine balloon tamponade versus uterine artery embolization in the management of persistent postpartum hemorrhage

Abstract

Objective

The aim of this study was to compare the cost of a strategy initially resorting to intrauterine balloon tamponade versus a strategy initially resorting to uterine artery embolization in the management pathway of persistent postpartum hemorrhage.

Methods

This was a propensity score–matched cohort study including women who were subsequently treated with (scenario 1) versus uterine artery embolization (scenario 2) in order to control PPH, defined as postpartum bleeding that was resistant to first‐line treatment.

We compared these two scenarios in terms of cost. These were based on costs of intervention, blood products transfused, and intensive care unit admission. We calculated the mean cost difference between the scenarios per woman.

Results

The propensity score–matched cohort comprised 50 women per scenario. In scenario 1, intrauterine balloon tamponade was successful in 29 women (58%), with a mean cost per patient of €7060 (standard deviation [SD], €2846). In scenario 2, uterine artery embolization was successful in 42 women (84%), with mean cost per patient of €7122 (SD, €2918). In each scenario, six women (12%) underwent a peripartum hysterectomy. The difference in mean cost per woman between the two scenarios was €62 in favor of scenario 1.

Conclusion

There is a negligible difference in cost between scenarios in which women with PPH are initially managed with intrauterine balloon tamponade versus uterine artery embolization. Given the comparable cost and maternal outcomes, intrauterine balloon tamponade is a favorable treatment, considering its less invasive nature, practical considerations, and lower risk of complications.

1. INTRODUCTION

Postpartum hemorrhage, the leading cause of maternal morbidity and mortality worldwide, poses a significant challenge in obstetric care. ¹ Most women respond well to first‐line treatment, depending on the cause of hemorrhage. For those with persistent postpartum hemorrhage (PPH), a peripartum hysterectomy can be lifesaving. ² , ³ , ⁴ A multidisciplinary approach is important in managing such cases. ⁵ Several less invasive interventions are available to manage ongoing bleeding refractory to first‐line treatment. These include intrauterine balloon tamponade (IUBT), uterine artery embolization (UAE) or ligation, and uterine compression sutures. ⁶ , ⁷ , ⁸ , ⁹

Intrauterine balloon tamponade has been studied and proven to be a promising intervention for managing PPH. ¹⁰ This intervention is often perceived as less resource‐intensive and is favored for its relatively easy application and minimally invasive nature. However, it is not without potential drawbacks, such as the risk of prolonged or more severe bleeding, and it could delay additional interventions, such as uterine artery embolization or ligation. Uterine artery embolization offers an alternative approach with higher reported success rates, defined by the reduced need for additional interventions. Nevertheless, it carries higher risks of complications and may not be available in all settings due to the required expertise and resources. ¹¹ , ¹²

In a previous report, we compared outcomes of women with PPH initially treated with intrauterine balloon tamponade versus uterine artery embolization, showing a success rate of 58% for intrauterine balloon tamponade and a success rate of 84% for uterine artery embolization, but with comparable risk of peripartum hysterectomy, number of red blood cell (RBC) units transfused, estimated blood loss (EBL), and maternal death. ¹¹ Supporting these findings, two French observational studies have reported a lower proportion of invasive interventions following the implementation of intrauterine balloon tamponade in their clinical practice. ¹³ , ¹⁴

Given the similar outcomes observed with both intrauterine balloon tamponade and uterine artery embolization, other factors such as availability, costs, and risk of complications become critical in choosing the most appropriate initial intervention for managing PPH. Therefore, to aid decision‐making in daily clinical practice, we aimed to assess the financial cost associated with intrauterine balloon tamponade versus UAT as treatment strategies for PPH.

2. MATERIALS AND METHODS

2.1. Study design and population

For the present study, we used data from the TeMpOH‐1 (Transfusion Strategies in Women During Major Obstetric Hemorrhage) study. ¹¹ This was a retrospective cohort study comprising consecutive women from 61 hospitals in the Netherlands who received at least four units of RBCs or a multicomponent blood transfusion within 24 hours following birth because of postpartum hemorrhage (≥1000 mL blood loss). Eligible women were selected from transfusion databases and birth registries of participating hospitals. From this cohort, we identified women with PPH, defined as postpartum hemorrhage with at least 1000 mL of blood loss refractory to first‐line interventions aimed at controlling bleeding (Table A1). ¹⁵

We used the propensity score–matched cohort as previously described. ¹¹ Due to propensity score matching, this cohort consisted of two groups of women with similar severities of PPH: those who were subsequently treated with intrauterine balloon tamponade versus those treated with uterine artery embolization. ¹¹ , ¹⁵ Figure 1 shows the matching procedure and the final cohort, with 50 women with PPH in both groups. For patient characteristics see Table A2. Women within the TeMpOH‐1 cohort were included in the current analysis if the treatment was assigned between an EBL of 1000 mL to 7000 mL, because there were no women with an EBL >7000 mL who received intrauterine balloon tamponade and there were no women with an EBL <1000 mL who received uterine artery embolization. Women were matched in the propensity score model on the following covariates: maternal age at time of birth, gestational age, parity, pre‐eclampsia, prior cesarean delivery, mode of birth, cause of hemorrhage, coagulopathy, symptoms of shock, EBL at the time of intervention, hemostatic interventions, administration of nonuterotonic agents, blood products transfused, and other surgical interventions.

FIGURE 1.

Study population and propensity score matching. EBL, estimated blood loss; IUBT, intrauterine balloon tamponade; PAS, placenta accreta spectrum; PPH, postpartum hemorrhage; RBC, red blood cell; TeMpOH‐1, Transfusion Strategies in Women During Major Obstetric Hemorrhage; UAE, uterine artery embolization.

2.2. Interventions

For this study we defined time zero as the moment during PPH when a clinician had to decide on what additional intervention to perform in order to stop the bleeding: intrauterine balloon tamponade or uterine artery embolization. During the study period, both interventions were part of the treatment algorithm of PPH in the Netherlands. At this point, we assumed that all women were in the operating theater for removal of the placenta or exploration of the uterine cavity for placental remnants after vaginal birth, or for cesarean section. We also assumed that all women were under general anesthesia. During the study period, the Bakri balloon (Cook Medical) was the type of intrauterine balloon device mostly used in the Netherlands. Intrauterine balloon tamponade was defined as the insertion of a intrauterine balloon catheter into the uterine cavity for the purpose of tamponade, which was performed in the operating theater. For intrauterine balloon tamponade, we assumed 30 min of additional operation time. For uterine artery embolization, we assumed women were transferred to an angiography suite, with obstetric and anesthetic teams present, and with a procedure time of 60 min. For laparotomy with hysterectomy, we assumed 120 min of operation time and for laparotomy without hysterectomy we assumed a procedure time of 45 min. Previous interventions were taken into account in the propensity score matching and were therefore not a part of the present analysis.

2.3. Cost analysis

For this cohort, we aimed to show the different sequential treatments women received over the course of PPH until the cessation of bleeding or death. We presented two scenarios: women who were initially treated with intrauterine balloon tamponade (scenario 1) or women who were initially treated with uterine artery embolization (scenario 2). The following treatments were presented: intrauterine balloon tamponade, uterine artery embolization, and laparotomy with or without hysterectomy. Per scenario we presented the following: (1) treatment success rate, (2) maternal outcomes of the women who were successfully treated by the specific treatment, (3) complications, and (4) cost of both scenarios.

2.3.1. Treatment success

We presented the treatment success rate for each intervention, along with any subsequent treatments and their outcomes if the initial intervention failed. Treatment success was defined as no need for additional surgical or interventional radiological procedures. For each treatment we provided the outcomes of the women who were successfully treated by the specific treatment. Additional considerations are reported in Appendix A2 (Supplemental Methods).

2.3.2. Maternal outcomes

For each intervention, we reported the following outcomes of the women who were successfully treated by the specific treatment: time until cessation of bleeding, total blood loss, number of blood products transfused, hemorrhagic shock, the need for intensive care unit (ICU) admission, and maternal mortality. The time until cessation of bleeding was defined as time, in minutes, between childbirth and last measurement of blood loss or last obstetric intervention if the last blood loss measurement was before the time of last intervention. The total blood loss was defined as the sum of intraoperative and postoperative blood loss in milliliters. We reported the following blood products transfusions: the number of RBC units, fresh frozen plasma units, and platelet concentrates transfused. Hemorrhagic shock was defined as at least one measurement of systolic blood pressure ≤90 mm Hg and/or heart rate ≥ 120 beats per minute during bleeding. Maternal mortality was defined as death of a woman while pregnant or within 42 days of termination of pregnancy. ¹⁶ We used the CROWN (Core Outcomes in Women's Health) initiative for treatment of postpartum hemorrhage to define the outcomes of interest in our study. ¹⁷

2.3.3. Complications

We aimed to report the number of complications per intervention, such as thromboembolic events, postembolization syndrome, uterine necrosis, infection, or uterine perforation. We did not have sufficient data on complications of the interventions in the TeMpOH‐1 data set. Therefore, we used the literature to present a crude estimate on risk of complications: 2.3%–6.5% for intrauterine balloon tamponade and 4%–18% for uterine artery embolization (Appendix A.5). ¹⁰ , ¹²

2.3.4. Costs

To calculate the costs per intervention, we aggregated costs associated with the utilization of operation rooms or angiography suites, personnel costs, and materials used. The costs of the operating room and angiography suite were based on the literature, ¹⁸ , ¹⁹ personnel costs were obtained from the Dutch manual for costing research, ²⁰ and material costs were based on purchase prices of the Leiden University Medical Center. Costs, reported in Euros, were adjusted to the year 2024 by the Dutch consumer price index. ²⁰ Specifications of the costs of all interventions are presented in Appendix B (Table B2) and explained in more detail in Appendix A.3.

In addition, we calculated the costs of the following maternal outcomes: transfusion of blood products, based on purchase prices of the Leiden University Medical Center, and ICU admission. There were no data available on the duration of ICU admission. Therefore, we assumed all admissions were ≤24 h, and, from a cost perspective, we used the costs for 24‐h admission in the ICU from the Dutch manual for costing research. ²⁰

2.4. Analysis

We presented descriptive statistics of the women in our propensity matched cohort study. We did not perform statistical tests. We presented continuous variables as medians with interquartile ranges (IQRs), and continuous variables as frequencies with percentages. We presented the cost difference between scenario 1 and scenario 2. We used means to calculate the mean cost per patient. We presented the cost per patient per scenario in a separate figure. To calculate the price per patient over the course of treatment, we divided both scenarios in subdivisions, as explained in more detail in the Supplemental Methods (Appendix A.4). We conducted our analysis using Stata version 16.1 (StataCorp LLC).

2.5. Ethical approval

Approval of the TeMpOH‐1 study and a waiver of informed consent was obtained from the medical ethics research committee of Leiden University Medical Center (reference number P12.273), and by the institutional review boards of all participating hospitals. The study was registered in the Dutch Trial Register (NL3909).

3. RESULTS

3.1. Treatment success

Within the propensity score–matched cohort, 50 women underwent treatment with intrauterine balloon tamponade (scenario 1), which was successful in 29 women (58%). Among women in scenario 1, subsequent uterine artery embolization after balloon tamponade was successful in 87% (13 of 15) and 12% (6 of 50) underwent hysterectomy, including two after embolization. Four women underwent hysterectomy directly after balloon tamponade, with two in hospitals without embolization availability. The other 50 women were initially treated with uterine artery embolization (scenario 2). In 84% of these women (42 of 50), uterine artery embolization was successful in cessation of bleeding. One patient underwent a second uterine artery embolization, which failed to stop the bleeding. Six women (12%) underwent a hysterectomy after uterine artery embolization. In both groups, eight women (16%) underwent a laparotomy, of whom six underwent a hysterectomy to stop bleeding. Two women in each group underwent an additional intervention during laparotomy, which was successful in cessation of the bleeding.

3.2. Cost

The mean cost per patient for scenario 1 was €7060 (standard deviation [SD], €2846) and for scenario 2 was €7122 (SD, €2918), with a mean difference of €62 in favor of scenario 1 (Figure 2, detailed description of the schematic representation of cost analysis is provided in Appendix C). The estimated costs of the interventions of interest were as follows: intrauterine balloon tamponade €645, uterine artery embolization €1656, laparotomy €780, and hysterectomy €1821. In scenario 1, the mean costs per patient of interventions (intrauterine balloon tamponade, uterine artery embolization, laparotomy, hysterectomy) were €1391, blood products transfused €4051, and ICU admission €1618. In scenario 2, the mean costs of interventions were €1939, blood products transfused €3855, and ICU admission €1329. Figure 3 presents the mean cost per patient per scenario. The mean cost for women who were successfully treated with intrauterine balloon tamponade was €4962 and for women successfully treated with uterine artery embolization was €5966. See Appendix D for our model to calculate costs.

FIGURE 2.

Schematic representation of cost analysis. ICU, intensive care unit; IUBT, intrauterine balloon tamponade; PPH, postpartum hemorrhage; RBC, red blood cell; UAE, uterine artery embolization.

FIGURE 3.

Cost per patient per scenario. Prices represent the mean cost per patient for a specific scenario. Costs are calculated as follows: € interventions + € blood products + € ICU admission. IUBT, intrauterine balloon tamponade; UAE, uterine artery embolization. Please find maternal outcomes used to calculate cost per patient per scenario in Appendix Table B3.

3.3. Maternal outcomes

In scenario 1, median blood loss was 4200 mL (IQR, 3500–4800 mL) for women successfully treated with intrauterine balloon tamponade (n = 29). For those who underwent intrauterine balloon tamponade followed by successful uterine artery embolization (n = 13), median blood loss was 5000 mL (IQR, 4300–6000 mL). In scenario 2, median blood loss was 4000 mL (IQR, 3300–4500 mL) for women successfully treated with uterine artery embolization (n = 42).

In women who underwent hysterectomy, median blood loss was 8650 mL (IQR, 7000–12 000 mL) in scenario 1 and 8000 mL (IQR, 7000–14 500 mL) in scenario 2 (Figure 2 and Table 1).

TABLE 1.

Maternal outcomes propensity score–matched cohort.

No additional intervention required, n (success rate, %)	Propensity score–matched cohort
	intrauterine balloon tamponade, n = 50				uterine artery embolization, n = 50
	Intrauterine balloon tamponade n=29 (58%)	Uterine artery embolization after intrauterine balloon tamponade n=13 (87%)	Laparotomy n=2	Hysterectomy n=6	Uterine artery embolization n=42 (84%)	Laparotomy n=2	Hysterectomy n=6
Time until stop bleeding (min) a	310 (165–411)	273 (260–339)	278, 285	456 (293–792)	361 (255–530)	525, 1541	644 (602–1612)
Total blood loss (mL) a	4200 (3500–4800)	5000 (4300–6000)	5400; 3800	8650 (7000–12 000)	4000 (3300–4500)	3500; 5500	8000 (7000–14 500)
Blood transfusions a
RBC transfusion (units)	6.0 (5.0–8.0)	9.0 (6.0–10.0)	11, 4	13.0 (11.0–17.0)	5.5 (4.0–8.0)	10, 8	17.0 (13.0–21.0)
Fresh frozen plasma (units)	2.0 (2.0–4.0)	6.0 (4.0–8.0)	3, 0	7.5 (5.0–13.0)	2.5 (2.0–5.0)	3, 0	8.0 (5.0–11.0)
Platelet transfusion (units)	0.0 (0.0–1.0)	2.0 (0.0–2.0)	1, 0	1.5 (0.0–3.0)	0.0 (0.0–1.0)	0, 1	2.0 (2.0–3.0)
Shock, n (%)	24 (82.8%)	10 (76.9%)	2	5 (83.3%)	35 (81.0%)	2	5 (83.3%)
Missing	1	1	0	0	1	0	1
University hospital, n (%)	3 (10.3%)	5 (38.5%)	0	2 (33.3%)	12 (27.3%)	0	2 (33.3%)
ICU admission, n (%)	13 (44.8%)	9 (69.2%)	1	5 (83.3%)	17 (40.5%)	1	5 (83.3%)
ICU admission, n (%)	28 (56%)				23 (46%)
Maternal deaths, n (%)	0				0

Abbreviations: ICU, intensive care unit; IUBT, intrauterine balloon tamponade; RBC, red blood cell; UAE, uterine artery embolization.

^a Median (interquartile range).

The variation in time until cessation of bleeding per scenario is shown in Figure 2 and Table 1. The number of RBC units transfused was comparable between women successfully treated with intrauterine balloon tamponade versus uterine artery embolization. In women with intrauterine balloon tamponade followed by successful uterine artery embolization (n = 13), the median number of RBC units transfused was 9 (IQR, 6–10). Results of all blood component transfusions are presented in Figure 2 and Table 1. Twenty‐eight (56%) women in scenario 1 were admitted to the ICU, compared with 23 women (46%) in scenario 2. There were no maternal deaths. See appendix Table B3 for maternal outcomes per scenario per patient.

4. DISCUSSION

The cost analysis presented indicates that in the Netherlands, there is no notable cost difference between the scenarios where women were initially treated with either intrauterine balloon tamponade or uterine artery embolization for managing PPH.

Our previous study has shown that intrauterine balloon tamponade has the potential to eliminate the necessity for embolization in managing PPH in most women, without increasing the risk of severe maternal outcomes. ¹¹ Previous studies, yielding conflicting results, highlight the existing variability in the application of intrauterine balloon tamponade. Variations in the choice of devices, such as balloon catheters or condom catheters, as well as differences in the clinical settings—whether in the labor room after vaginal birth or in the operating room—contribute to the heterogeneity observed in the literature. ⁹ , ¹⁰ , ²¹ , ²²

This cost analysis did not show cost benefits for women with PPH who were initially treated with intrauterine balloon tamponade. With our propensity score–matched cohort, we have a selected population of women with severe postpartum hemorrhage in which a clinician was willing to perform uterine artery embolization. This group represents the most severe cases within the broader population of women with PPH. In our previous study, the total cohort consisted of 455 women with an EBL between 1000 and 7000 mL at the time of treatment assignment, of whom 373 were initially managed by intrauterine balloon tamponade and 82 with uterine artery embolization. We reported an intrauterine balloon tamponade success rate of 70% in the total group from the unadjusted analysis, which is in line with previous literature. ¹⁰ , ¹¹ The higher success rates in the literature and in our unadjusted analysis could relate to the difference in severity of bleeding. In our present cost analysis, we observed the costs of transfusion of blood products to be the largest proportion of the total costs. The mean number of blood products transfused likely varies according to the severity of bleeding. In general, early timing of intrauterine balloon tamponade could not only improve maternal outcomes but may also be beneficial from a cost perspective. The current cost analysis suggests that, even in severe cases of PPH, intrauterine balloon tamponade is an appropriate treatment, as there is no difference between cost of intrauterine balloon tamponade and uterine artery embolization.

Factors that could advocate for intrauterine balloon tamponade as an appropriate choice in the management of PPH include its lower risk of complications, wide availability, and minimal need for specialized training or expertise. Intrauterine balloon tamponade can be easily administered in various clinical settings, including in high‐ and low‐resource settings. ² , ²³ The procedure's simplicity ensures that it can be effectively performed by clinicians without requiring extensive training, in contrast to more complex procedures such as uterine artery embolization. Additionally, studies have demonstrated a higher risk of complications associated with uterine artery embolization compared with intrauterine balloon tamponade. ¹⁰ , ¹² Given these advantages, intrauterine balloon tamponade presents as a pragmatic and effective treatment option for managing PPH.

Our study has some limitations. Due to the necessity of making various assumptions to conduct the cost analysis, our results represent a rough estimation. Costs were calculated for a university hospital in the Netherlands and might not be representative of other settings. For this cost analysis, we adopted a healthcare perspective and did not evaluate associated nonhealthcare costs. The societal perspective, including also costs related to the time to return to work, was not part of our analysis. A comprehensive cost‐effectiveness analysis would also require the inclusion of quality of life as an outcome measure. Since maternal outcomes were similar between both scenarios, we assume that the time to return to work and quality of life might also be comparable in both groups. ²³ We did not have sufficient data on complications for either intervention in the TeMpOH‐1 data set, and therefore the cost of complications are not considered in the cost analysis. We assumed a setting in which each hospital was equipped to perform either intrauterine balloon tamponade or uterine artery embolization. Our findings are not generalizable to a setting with different availability of interventional radiology. However, one of the strengths of our study is that, to the best of our knowledge, it is the first cost analysis in this area. Our cohort consists of women with varying causes of hemorrhage, predominantly PPH due to uterine atony. In total, 12 women underwent peripartum hysterectomy, six of whom had placenta accreta spectrum disorder, evenly distributed between both groups. These distributions indicate that the severity and underlying causes of hemorrhage were comparable between the groups, as a result of the propensity score model. Nonetheless, the inclusion of these severe cases in our cohort should be considered before generalizing our findings.

5. CONCLUSIONS

There is a negligible difference in cost between the scenarios in which women with PPH are initially managed by intrauterine balloon tamponade or by uterine artery embolization. Given the comparable cost and maternal outcomes, intrauterine balloon tamponade is a favorable treatment, considering its less invasive nature, practical considerations, and lower risk of complications.

AUTHOR CONTRIBUTIONS

Concept, design, and obtained funding: LB, PR, CCD, CG, DH, JvdB, TvdA, and EvdA. Analysis: LB, PR, CCD, DH, JvdB, TvdA, and EvdA. Drafting of the manuscript: LB, DH, JvdB, TvdA, and EvdA. Interpretation of data and critical review of the manuscript: all authors.

FUNDING INFORMATION

This work was supported by an internal grant from Sanquin Center for Clinical Transfusion Research (grant number PPOC‐11‐032).

CONFLICT OF INTEREST STATEMENT

The authors have no conflicts of interest.

Supporting information

Data S1.

Bonsen LR, van den Akker‐van Marle ME, Caram‐Deelder C, et al. Cost analysis of intrauterine balloon tamponade versus uterine artery embolization in the management of persistent postpartum hemorrhage. Int J Gynecol Obstet. 2025;170:1337‐1345. doi: 10.1002/ijgo.70149

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.