Postpartum hemorrhage care bundles to improve adherence to guidelines: A WHO technical consultation

Abstract Objective To systematically develop evidence‐based bundles for care of postpartum hemorrhage (PPH). Methods An international technical consultation was conducted in 2017 to develop draft bundles of clinical interventions for PPH taken from the WHO's 2012 and 2017 PPH recommendations and based on the validated “GRADE Evidence‐to‐Decision” framework. Twenty‐three global maternal‐health experts participated in the development process, which was informed by a systematic literature search on bundle definitions, designs, and implementation experiences. Over a 6‐month period, the expert panel met online and via teleconferences, culminating in a 2‐day in‐person meeting. Results The consultation led to the definition of two care bundles for facility implementation. The “first response to PPH bundle” comprises uterotonics, isotonic crystalloids, tranexamic acid, and uterine massage. The “response to refractory PPH bundle” comprises compressive measures (aortic or bimanual uterine compression), the non‐pneumatic antishock garment, and intrauterine balloon tamponade (IBT). Advocacy, training, teamwork, communication, and use of best clinical practices were defined as PPH bundle supporting elements. Conclusion For the first response bundle, further research should assess its feasibility, acceptability, and effectiveness; and identify optimal implementation strategies. For the response to refractory bundle, further research should address pending controversies, including the operational definition of refractory PPH and effectiveness of IBT devices.


| INTRODUCTION
Postpartum hemorrhage (PPH) occurs in approximately 5% of all live births and, despite concentrated efforts, remains a leading cause of maternal morbidity and mortality. 1 Because most PPH-related deaths are preventable through the implementation of effective interventions, the recent shift from home births to facility births across lowand middle-income countries (LMIC) raises new opportunities for saving women's lives. 2,3 Unfortunately, inconsistent and/or delayed use of effective interventions for prevention and treatment of PPH, in addition to other systemic problems in health services (e.g., lack of blood banks, inadequate staffing), has led to continued unacceptable rates of hemorrhage-related maternal deaths. [4][5][6] Care bundles have been associated with improved patient outcomes when adherence is high. [7][8][9] The concept of care bundles is similar to that of packages and checklists, which have been used by healthcare providers for decades with a similar goal of standardizing and expediting care (Supplementary Box S1). Care bundles may include behaviors, such as the widely used "ABCs" designed to help practitioners remember the sequence for resuscitation, or a number of interventions packaged together, such as the "Active Management of the Third Stage of Labor" (AMTSL) package used to prevent PPH.
In 2001, the Institute for Healthcare Improvement (IHI) developed a formal approach to bundling care to increase the quality and efficiency of care delivery. 10 The IHI defined bundles as "small sets of evidence-based interventions for a defined patient population and care setting that, when implemented together, result in significantly better outcomes than when implemented individually". 10 The "bundles" approach was designed to increase uptake and compliance to recommended interventions. 10 Care bundles differ from other care packages in that compliance is achieved only when all the bundled interventions are completed and recorded. Thus, compliance for the bundle as a whole implies higher rates of compliance for its individual elements. 10 Teamwork, communication, and cooperation are emphasized, because these health systems' processes are required for quality and sustainability. 10 In 2012, WHO published its "Recommendations for the Prevention and Treatment of Postpartum Haemorrhage" to provide evidenceinformed clinical care recommendations for hemorrhage due to uterine atony. 11 However, adherence to these recommendations remains a challenge. 6 The bundle approach has been proposed as a potential PPH recommendations and based on the validated "GRADE Evidence-to-Decision" framework. Twenty-three global maternal-health experts participated in the development process, which was informed by a systematic literature search on bundle definitions, designs, and implementation experiences. Over a 6-month period, the expert panel met online and via teleconferences, culminating in a 2-day in-person meeting.

Results:
The consultation led to the definition of two care bundles for facility implementation. The "first response to PPH bundle" comprises uterotonics, isotonic crystalloids, tranexamic acid, and uterine massage. The "response to refractory PPH bundle" comprises compressive measures (aortic or bimanual uterine compression), the nonpneumatic antishock garment, and intrauterine balloon tamponade (IBT). Advocacy, training, teamwork, communication, and use of best clinical practices were defined as PPH bundle supporting elements.

Conclusion:
For the first response bundle, further research should assess its feasibility, acceptability, and effectiveness; and identify optimal implementation strategies. For the response to refractory bundle, further research should address pending controversies, including the operational definition of refractory PPH and effectiveness of IBT devices. solution to suboptimal adherence to PPH guidelines. 4 Healthcare bundles have been proposed for maternal conditions including placenta previa, elective induction, labor augmentation, vacuum delivery, maternal sepsis, and obstetric anal sphincter injury, 10,12,13 but evidence of their success or failure is lacking. Although many current patient safety programs target PPH, 3,5,[14][15][16] there are no patient care bundles for PPH as defined by the IHI.

K E Y W O R D S
In early 2017, WHO decided to explore whether bundling current WHO-recommended evidenced-based interventions for PPH due to uterine atony might accelerate adoption and adherence to PPH guidelines. The aim of the present study was to describe the first steps toward that goal: the adoption of a bundle definition, the PPH intervention selection criteria, and the process for the development of two PPH care bundles.

| MATERIALS AND METHODS
The consultation for the development of care bundles for PPH was carried out among international maternal health experts between Postpartum hemorrhage was defined as bleeding that a skilled birth attendant (SBA) feels is excessive and worrisome for this exercise. 17 In addition, in the absence of an accepted definition of refractory PPH, it was defined as bleeding that is unresponsive to initial treatment and that triggers an additional set of interventions.
Development of the bundles was undertaken by a panel of experts with geographic and professional diversity (Supplementary File S1).
The PPH bundles were developed first by conducting a systematic literature search to define care bundles and their essential characteristics in general; and then by identifying criteria to guide the selection of interventions for the PPH bundles. The selection of the interventions to be included in the bundles was made through technical consultations. Figure 1 outlines the process followed for bundle development.  (Table 1). For settings we considered community settings (i.e., home deliveries, health after delivery, and dispensary deliveries) assisted by SBAs, primary healthcare (PHC) centers, and hospitals. All WHO recommendations were assessed for appropriateness within each of these settings, resulting in 13 interventions eligible for inclusion ( Table 2).
The 13 interventions were then classified according to purpose (prevention, first response, and response to refractory PPH); setting (as above); application to vaginal delivery, cesarean delivery, or any type of delivery; and application during the third stage of labor or the first 24 hours postpartum. From a total of 38 possible combinations of the 13 interventions that emerged from the above classification, those that included three or more interventions, were judged to be applicable in most settings, were intended for use by skilled birth attendants, and would be applicable to most women with PPH due to uterine atony were selected. Two recommended interventions, hemostatic surgery and arterial embolization, were excluded from the bundles because neither is feasible in most settings nor applicable to most women with PPH due to uterine atony. The experts met for an in-person consultation December 7-8, 2017, to consolidate agreements and to address disagreements.
Presentations and discussions were held in plenary sessions, where the "poll everywhere" audience response system and paper ballots were used to record individual decisions anonymously. See Supplementary File S5 for details.

| RESULTS
In the literature search, 730 articles met the initial criteria, of which 415 were excluded after reviews of the abstract and full text  What is the certainty of the evidence for the costs of the intervention? If resource use is considered critical for a recommendation, the less certain the evidence for resource requirements, the less likely it is that a panel should select or not the intervention 6 (4-7.5) No 7 Cost-effectiveness Judgments about the cost effectiveness of an intervention need to take into account several criteria including the balance between the desirable and undesirable effects (the net benefit); the certainty of the evidence of effects and uncertainty about or variability in how much individuals value the main outcomes; and resource requirements (cost) and uncertainty about the costs 7.5 (7-8) Yes 8

Resources required
How large are the resource requirements (costs in terms of both money and time) of the bundle? The greater the cost, the less likely it is that a bundle will be selected 8 (7.5-9) Yes 9 Equity What would be the impact of the bundle on health equity? This criterion evaluates if a bundle is expected to reduce health inequities. It considers whether a bundle will reduce differences in the effectiveness for disadvantaged populations within countries, such as low-income groups, less educated individuals, and/or rural populations 8 (6.5-9) Yes 10 Acceptability Is the bundle acceptable to key stakeholders (women and providers)? A bundle might vary on its acceptability level due to ethical principles (e.g., autonomy, beneficence or justice), as well as the distribution of the desirable and undesirable effects and costs (who benefits or is harmed, and who pays or saves) 8 (6.5-9) Yes  Tables S1 and S2).

| Prevention and recognition of PPH bundle
The bundle of interventions proposed for PPH prevention included uterotonics, controlled cord traction (CCT), and uterine tone assessment. In the online rounds, this bundle received high relevance rates and strong agreement overall (Supplementary Table S3

| Response to refractory PPH bundle
The following set of interventions was proposed for the response to refractory PPH care bundle: continue administration of uterotonics and isotonic crystalloids, second dose of TXA, IBT, and non-pneumatic antishock garment (NASG). It was acknowledged that IBT or NASG may not be available in some settings.
During the online consultation, this bundle, intended for women who continue to bleed despite implementation of the first response bundle and whose condition worsens or deteriorates, received high RAND relevance scores, and had the agreement of the panel for the PHC and hospital levels. For the community level, however, the bundle received low RAND relevance scores for four criteria (acceptability, feasibility, indicator measurability, and no or minimal resources required), and there was no consensus for the equity criteria (Supplementary Table S3).
During the discussions at the in-person meeting, the following issues were discussed for the refractory bleeding bundle. First, uterotonics, crystalloids, and TXA were already included in the first response bundle, and therefore did not need to be listed as bundle components.
Second, IBT is currently recommended by the WHO, but is considered controversial by some members owing to recently published evidence. [22][23][24][25] Third, in cases where IBT or NASG is not available, or for use during the period before IBT and NASG are applied, bimanual uterine compression and external aortic compression were suggested for bundle inclusion by some experts. Fourth, concerns were raised that implementing all elements of the bundle might result in the overtreatment of women with refractory hemorrhage whose condition was not worsening. Last, an area of contention was whether or not the "response to refractory PPH bundle" should be a bundle. Some members mentioned that the conditional, variable, and progressive changes of refractory hemorrhage may make this condition less appropriate for the bundle approach.
In response to these concerns, the panel considered the follow- The primary rationale for keeping these interventions in a bundle was, first, that the "care package" approach has been recommended by WHO since 2012; and second, the rationale for proposing a bundle approach was to improve strategies for compliance with best practices.
The original aim was that the PPH bundles would apply to both bundles" and will be re-examined as new evidence emerges during the process of updating WHO recommendations and guidelines. 26

| DISCUSSION
In the present consultation, a systematic approach was used to review  supply chains and encouraging behavior change are critical to implementation. Assessment of facilitators and barriers should guide the development of the strategy. The approach will need to be tailored to local contexts to ensure sustainability. Similarly, leadership from ministries and key stakeholders will be critical for successful bundle implementation. We expect that the PPH bundles will reduce rates of severe PPH, morbidity, and mortality, through improved quality of care and adherence to global, high-quality guidelines; however, this has not yet been demonstrated. Future research must rigorously assess how these bundles are implemented in practice, including the mechanisms of impact and how these are influenced by the context. 28 [Correction added on 21 January 2020, after first online publication: the reference citation 29 was changed to 28.] Factors to be evaluated include bundle feasibility, acceptability, safety, adverse consequences, and effectiveness relative to individual interventions.
The opinions of healthcare planners, practitioners, and users will be important to consider. Cost-effectiveness and impact should be stud- Given these considerations, there will be a need for implementation research to determine if the bundling approach will ultimately make a difference in saving women's lives from PPH.

Supporting Information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Figure S1. Flowchart of the systematic search of the literature. Table S1. Type of bundles and interventions based on the systematic search of the literature.  File S5. Technical consultation agenda.
Box S1. Examples of best clinical practices.