Populations.

• Women of any parity (including nulliparous) who gave birth in any setting (including in a health facility or at home).
• Birth attendants who care for women during childbirth including healthcare providers, nurses, midwives, doctors, skilled birth attendants, lay health workers, community health workers or traditional birth attendants (see search terms in S1 Appendix for full list).

Exposure.

• Third stage of labour including a range of interventions and procedures performed for the prevention and treatment of post-partum haemorrhage

Outcome.

• The views, experiences and perceptions of the populations under investigation with particular regard to the acceptability and feasibility of approaches relating to clinical practices for PPH prevention.

The search terms were organized into broad strings to cover the PEO items outlined above as well as a search string to cover qualitative methodologies. An example of the search strategy is shown in S1 Appendix. In order to achieve an optimal number of studies eight relevant databases were searched, Medline [OVID]; CINAHL [EBSCO]; PsycINFO [EBSCO]; EMBASE [OVID]; Global Index Medicus; LILACS (for studies conducted in South America); AJOL (for studies conducted in Africa) and POPLINE. The searches were conducted on 9^th and 10^th July 2018.

Where possible, research limiters were used to ensure that searches for qualitative studies were optimized. For example, the research limiter for MEDLINE located ‘Qualitative studies–best balance of specificity and sensitivity’. We searched the reference lists of all eligible studies and used back chaining to locate any relevant studies that were not identified in the electronic searches.

The review was used to inform the updating of WHO’s recommendations on prevention and treatment of PPH, so our searches also included terms relating to the treatment of PPH (e.g. balloon tamponade, NASG, etc;). This paper reports on those studies focusing on the prevention of PPH and a subsequent paper will report on the views and experiences of relevant stakeholders with regard to the treatment of PPH.

Intervention(s).

We were interested in the prevention of PPH from a qualitative perspective so were looking for any qualitative studies that explored this phenomenon from the perspective of women and healthcare providers. Preventative approaches included the use of uterotonics (oxytocin, carbetocin, ergot alkaloids, ergotamine, ergometrine, oxytocin-ergometrine, misoprostol or prostaglandin) or any associated procedures or practices that influenced the prevention of PPH (active management, uterine massage, uterine packing, uterine compression, cord clamping, controlled cord traction, retained placenta, placental delivery).

Study selection.

We collated records identified from the different sources into a data management programme (RefWorks) and removed any duplicates. One review author (KF) assessed each abstract in order to determine eligibility for inclusion against the a priori inclusion/exclusion criteria. At this stage we removed any abstracts that were clearly irrelevant to the focus of the review. Two review authors (KF, SD) then retrieved the full texts of all the papers that were likely to be relevant, and assessed them independently for eligibility, before agreeing on the final list of included studies. In the event of lack of consensus over the inclusion of a particular study, a third review author (OTO) was available to adjudicate.

We recorded study characteristics using a form designed specifically for this review. The form includes details of: study author, date of publication, country of study, context (urban/rural), region (Africa, Americas, South-East Asian, European, Eastern Mediterranean, Western Pacific), participant group (women or healthcare providers), theory and method used, sample size and quality grading.

Quality assessment.

The included studies were subject to quality appraisal using an instrument developed by Walsh and Downe [19] and modified by Downe et al [21]. This is a simple appraisal system based on an amalgamation of several published qualitative appraisal tools including the Critical Appraisal Skills Programme (CASP). The system rates studies against 11 pre-defined criteria, and then allocates a score from A-D (see below). Studies scoring C+ or higher will be included in the final analysis.

Scoring criteria for quality appraisal

A: No, or few flaws. The study credibility, transferability, dependability and confirmability are high;

B: Some flaws, unlikely to affect the credibility, transferability, dependability and/or confirmability of the study;

C: Some flaws that may affect the credibility, transferability, dependability and/or confirmability of the study.

D: Significant flaws that are very likely to affect the credibility, transferability, dependability and/or confirmability of the study.

Data extraction and analysis.

Data extraction and analysis were conducted simultaneously in four steps using methods aligned to meta-ethnography [18] to produce descriptive Summaries of Findings (SoFs–second order constructs). The Summary of Findings statements were evaluated for confidence using the GRADE-CERQual tool (Grading of Recommendations, Assessment, Development and Evaluation)-(Confidence in the Evidence from Reviews of Qualitative Research) [20]. They were then synthesised into third order themes as a basis for deriving a line of argument synthesis from the data.

In step one, basic details of the included papers (author, date, title, country) were indexed and organized into two categories–those representing the views and experiences of women and those representing the views and experiences of health care professionals. For each category the papers were examined, and an index paper selected, chosen to best reflect the focus of the review. The themes and findings identified by the authors of this paper were entered onto a spreadsheet. This process was then continued for all the remaining papers, one at a time, iteratively building a set of themes that captured the quote material presented by the original authors, and their themes and statements, forming the ‘first order constructs’ of this stage of the review.

In step two the first order constructs were refined and merged to form second order constructs (Summary of Findings [SoF] statements), at one remove from the actual data in the included studies. This process includes looking for what is similar between papers (‘reciprocal analysis’), and for what contradicted the emerging findings (‘refutational analysis’). The disconfirming data identified in the on-going refutational analysis were used to refine the emerging constructs, so that the final analysis had high explanatory power for all the data. The second order constructs were developed by KF and agreed by consensus between KF and SD.

In step three, the second order constructs (the SoFs) were assessed for confidence using the GRADE-CERQual tool. This is a recently developed instrument, derived from the approach used in quantitative effectiveness reviews (GRADE). CERQual’s assessment of confidence for individual review findings from qualitative evidence syntheses is based on four components: the methodological limitations of the qualitative studies contributing to a review finding, the relevance to the review question of the studies contributing to a review finding, the coherence of the review finding, and the adequacy of data supporting a review finding. Based on these criteria, review findings were graded for confidence using a classification system ranging from ‘high’ to ‘moderate’ to ‘low’ to ‘very low’. As with study selection, the grades for each review finding were agreed by consensus and where there was disagreement a third reviewer (OTO) was asked to arbitrate.

In step four, the graded review findings were collapsed into over-arching interpretive themes as a means of synthesizing the data into an underlying ‘line of argument’ that describes the whole data set succinctly.

Understanding of blood loss.

For women living in a variety of predominantly rural settings in LMIC’s the release of 'dirty blood' after childbirth was perceived as a normal cleansing process and something that shouldn't be prevented. “The blood that flows after childbirth is bad blood because for the whole pregnancy you are not menstruating and this is a long period. When that blood remains in your womb, you might get some complications”. (20 year old woman, Uganda [29]). In communities where this view was prevalent it was held by both childbearing women and traditional birth attendants (TBA’s), who saw little reason to adopt preventative measures and sometimes regarded prevention of postpartum blood flow as a dangerous process. The distinction between what constituted normal blood loss and what might be considered dangerous was often blurred and relied on simplistic measures such as visual inspection of buckets, cloths and pads. However, these respondents also showed awareness of the physiological signs of extreme blood loss and shock: The woman’s body becomes blue (nila hoiya jay) from continuous bleeding. The blood goes out in a continuous, swift flow or gush (dhala dhala), it overflows the place. Two or three jute-made bags (chala-chula) will not control or stop the blood. (Postpartum woman, Bangladesh [30]).

Impact of PPH.

Women in several high-income settings recounted their experience of having a PPH and their hope that their stories might raise awareness of the condition from a preventative perspective. They talked about overwhelming feelings of exhaustion and fatigue following the PPH as well as powerful feelings of disorientation that affected communication with healthcare staff during the PPH. They reported feeling anxious and frustrated at not being able to ‘bond’ with their baby immediately after birth and discussed the long-term psychological impact of having a traumatic experience, “More (information) is needed about the emotional difficulties after experiencing such a traumatic birth. I wasn’t prepared for this and suffered as a result” (Primiparous woman, 2 months postpartum, New Zealand [27])

Traditional means of preventing or treating PPH.

In a number of different LMIC's women adopted a variety of traditional practices to prevent and treat PPH. In some contexts spiritual or supernatural forces were thought to be responsible, and prevention therefore depended on appeasing these forces: “Dushi (evil spirit) turns on a pregnant woman if she stays outdoors in the early morning, in the evening, and at noon, ignoring rules, and heavy bleeding occurs in this case” (Woman in Bangladesh [28]). Prevention of PPH also included mechanical techniques that recognised the need for compression: ‘Immediately after the expulsion of the placenta, I massage the woman’s abdomen, tie a cloth around the abdomen and advise her to lie on her tummy until the bleeding stops.’ (56 year old female TBA, South Africa [50]). In some contexts TBA’s used herbs and teas with uterotonic properties to prevent PPH. Though such treatments might actually have pharmacological justification, the practice was frowned upon by some healthcare professionals: “We do not know what the side effects might be. The problem is that the matrone (TBA) does not even know about the right dosage, what may happen if it is too much…For example, she may tell the woman to boil a handful of the plant but the problem is that a handful for the matrone might be different from the handful of the woman. So, there might be a risk of overdose and I am sure that it provokes [retained placenta]. (Obstetrician, Madagascar [47]).

Knowledge about PPH.

Several studies conducted in LMIC’s showed that women were often knowledgeable about PPH. Younger women, in particular, acknowledged that it was a complication that could and should be prevented. Where explanations were given they were generally vivid and clear: “This bleeding is so severe that a pad or cloth cannot hold it. The bleeding runs down the legs of the woman. At times it is like a running tap that the bleeding can fill up a bucket. Such bleeding may cause the death of the woman” (Woman in rural Nigeria [33]).

Desire for more information.

Although knowledge of PPH as a potential complication was fairly widespread, the desire for more information about all aspects of the condition was expressed by women in a number of settings, especially in high-income countries (HIC’s). Participants reflecting on their experiences of a PPH described poor communication by healthcare providers as a particularly frustrating aspect of their care and suggested information leaflets and/or relevant websites might address some of their concerns, especially if these were discussed with healthcare providers during antenatal appointments.

Desire for more staff training.

Across a wide range of settings and contexts, healthcare providers identified knowledge gaps in their understanding of PPH management, expressed a lack of confidence in their ability to manage PPH and highlighted a lack of training opportunities to address these concerns, particularly in LMIC's. “I have been working for over ten years and have had a refresher course only once. You can imagine, every time they (policy makers) talk about maternal death instead of giving health workers some more training on PPH….” (Medical assistant, Malawi [43]). In high-income settings the uncertainty around AMTSL practices caused confusion amongst health professionals and was usually attributed to lack of knowledge “Even if it is stated in the (PPH) protocols, team members still doubt the content, which leads to tedious discussions. This is mainly caused by lack of knowledge” (Junior doctor, The Netherlands [54]).

Provider preferences for PPH prevention.

Some healthcare providers remained unconvinced about the benefits of using uterotonics (primarily oxytocin) to prevent PPH and ignored pertinent guideline recommendations to better align with their personal beliefs. These beliefs tended to be based on notions of normality and the understanding that a woman’s embodied experience would lead to the appropriate course of action, “I don’t completely agree with having guidelines that must be strictly followed without thinking why; when you give it, why do you give it? So I choose the middle way, I mean if a normal primipara isn’t bleeding I write ‘Not given oxytocin’ in her notes. I see no reason to give it” (Midwife, Sweden [38]).

Health professionals also expressed a preference for a 'hands off' approach to third stage management believing that this technique didn't add to women's pain and produced better PPH outcomes when birth was perceived to be normal. “After the delivery I didn’t give oxytocin because she delivered normally. Only if she is bleeding. Because some of them they deliver normally, no problem.” (Maternal and Child Health Aide, Sierra Leone [49]).

Resource constraints.

Staff shortages and lack of medicines or equipment were consistently highlighted by healthcare providers in a wide range of settings and contexts. Staff shortages were particularly acute in a number of LMIC’s and were blamed for maternal deaths, “Another major cause of maternal death is lack of enough skilled workers in this midwifery practice” (Nurse-midwife, Malawi [43]). Inconsistent or limited stock levels of oxytocin were widespread in a number of LMIC’s and meant that providers were sometimes forced into using EMTSL, even though they perceived this to be sub-optimal care, “We give oxytocin if we have, but if we don’t have then we express normal procedures [expectant management].” (Maternal and Child Health Aide, Sierra Leone [49]). Limited supplies of oxytocin at health facilities sometimes had a knock-on effect for women because healthcare providers insisted on following AMTSL recommendations (out of fear of litigation) and asked women to buy oxytocin privately before coming to the health facility to give birth, “It’s a challenge when you tell them go and buy this drug because the hospital can't provide it and they discover that it’s expensive. The hospital provides it sometimes, but sometimes it runs out.” (Junior doctor, Uganda [45]).

Trust in Task Shifting.

To address the relatively high maternal mortality rates in some LMIC’s, national and local governments have introduced policies to train cadres with lower levels of training (mainly community health workers and traditional birth attendants) on how to use uterotonics, largely misoprostol, to prevent PPH in community settings. In some contexts, especially where home births were commonplace, health professionals thought this was a good idea, “I don’t differentiate between the health extension workers and the trained volunteers or the TBAs or anyone else. As long as they have received the proper training, they are more than capable of administering the drug”. (Healthcare provider, Ethiopia [57]). However, in other contexts, healthcare providers felt that TBA's and community health workers might use uterotonics irresponsibly, “You see they [TBA’s] know it as “Tha ein pwint say” and we are worried that they would use it before delivery thinking that it will enhance labour. So we emphasize in the training as “tha ein kyunte say” and we stress that it can only be given after the birth of the baby” (Auxiliary midwife trainer, Myanmar [53]).

Influence of hierarchical structures in perpetuating poor practice.

In a limited number of settings healthcare providers felt that their ability to influence change in the practice of AMTSL was hampered by existing hierarchical structures. A culture of repressive seniority coupled with a fear of repercussion prevented genuine dialogue that might improve clinical skills, “We were told to do the counter traction this way (showing the right manoeuvre), put pressure on the uterus upwards, and then start retracting the placenta. But when I came here, I realized some of my colleagues do it the other way round, they hold the uterus this way (showing the opposite grip) and they pull. But they are my seniors so if I say “Don’t do it this way or that way”, they would feel that I was looking down at them” (Labour ward midwife, Ghana [40]).

Importance of clinical competence.

Women in a limited number of high-income settings expressed a sense of reassurance when staff were perceived to be clinically competent. They felt safe when health professionals gave injections efficiently and painlessly and remained calm when AMTSL procedures were carried out with minimum fuss. By contrast, for a small number of women in rural Nigeria the perceived incompetence of staff at the local health facility deterred women from attending for potential childbirth complications. For these women the local TBA’s were regarded as more competent in PPH management. The author of this study notes that, “nurses and trained midwives are both ill-experienced and untrained to undertake simple methods of controlling haemorrhage, or handling other complications” [33].

Importance of woman-centred care.

As with most other aspects of maternity care the importance of kind, empathic, personal care was highly valued by women in a variety of settings. Being acknowledged and reassured were key features of this approach, “the care up to that point had been so good and everybody had been, um, so supportive, not that I’d expected otherwise but I was quite, um, pleased with the time they took just to reassure me and … that sort of personal touch, do you know what I mean, was really nice” (Post-natal woman, UK [24]). For some women the opportunity to engage with staff on a personal level demonstrated a woman-centred approach, while for others the simple act of receiving attention was perceived favourably, "The treatment was good, I liked it a lot. The nurse at sixth floor gave me a lot of attention and she continued in the delivery room" (Multiparous woman, Angola [31]).

Importance of shared-decision making.

Women and healthcare providers both discussed shared decision-making when talking about their experiences of AMTSL. Women expressed feelings of frustration and anxiety when decisions were taken without their involvement and feelings of control and empowerment when they were actively engaged in decisions relating to third stage practices, “I was very appreciative of the fact that [my midwives] gave me the info that I could handle for the situation that I was in. They didn’t hide anything, and they certainly answered every question that I had…so I was fine. I was calm” (Multiparous woman, Canada [25]). Healthcare providers in a limited number of settings also recognized the benefits of shared decision making and used terms like ‘option’ and ‘choice’ rather than ‘recommendation’ when discussing clinical practices in the third stage of labour [39].

Disrespect and abuse by staff.

In a relatively small number of diverse settings (including HIC’s and LMIC’s) women experienced disrespectful, rude and sometimes abusive behaviour by healthcare providers. In some LMIC contexts, there was an apparent expectation that women should ‘behave’ during all stages of childbirth and, if they failed to do so, they would be punished by healthcare providers, "I was well treated because I did not cause a scandal. The others who were shouting and crying were badly treated" (Multiparous woman, Angola [31]). In other low income contexts women experiencing a secondary PPH were punished, admonished, mistreated or even refused care because they had given birth at home rather than in a health facility, “No we did not go to the hospital [for treatment] because if they notice that you delivered from the village … she [the nurse] can easily beat you. So it is better I go to the old lady in the village because I know she will treat me well”. (Postpartum woman, Uganda [36]).

Recognition of the benefits of using uterotonics for PPH prevention.

Healthcare providers in a wide variety of settings and contexts acknowledged that both oxytocin and misoprostol were effective agents in strategies to reduce PPH, “One of the protocols is that when the woman has delivered we give oxytocin. . . . All of this management is an attempt to prevent PPH.” (Clinical Health Officer, Sierra Leone [49]). A number of healthcare providers had been involved with advance community distribution programmes and felt that misoprostol was a more practical and acceptable solution to PPH prevention in these contexts, “The women are grateful because they are afraid of bleeding. But with this tablet (misoprostol) they become more at ease. She was bleeding when she delivered her first baby, but the second one, she participated in the programme, got the medicine and there was no bleeding” (TBA, West Java [52])

Safety concerns about uterotonic use.

Healthcare providers in a number of LMIC’s highlighted specific safety concerns around the use of misoprostol, particularly in community settings. Although side effects were identified (especially shivering), they were usually considered to be transient and reasonably acceptable to women. Concerns were more commonly raised by providers when discussing advanced community distribution programmes and the potential for mis-use of the drug, “No one should overestimate misoprostol because many things can cause bleeding. What will happen if the drug is given when the woman is still in labor? Or if there is another baby (twin)? Or perhaps it may be used unsafely for other purposes… A woman can die”. (Healthcare provider, Ethiopia [57]). Some healthcare providers also expressed concerns about the administration of oxytocin and the perception that it might cause retained placenta “Some women are in a lot of pain and it’s my impression that we’ve had more retentions since we started giving those 10 units of oxytocin. The placenta gets trapped in the cervix” (Experienced midwife, Sweden [38]).

Implementation concerns.

In a number of LMIC’s healthcare providers had reservations about advance distribution programmes for misoprostol. For some policy level administrators, there were unanswered questions about the effectiveness of these programmes, “There is a general policy that guides us. But how should misoprostol be given? Who should it be given to? How it should be implemented? A document that includes all of these things should be compiled. I think there is much that has not been done”. (Healthcare provider, Ethiopia [57]). For others, there were concerns that advance distribution programmes (for misoprostol) could undermine efforts to strengthen safe delivery services or encourage unregulated providers to distribute misoprostol irresponsibly, “There needs to be controls otherwise sellers will advertise and people will misuse [the drug]” (Healthcare provider, Laos [56]). In a smaller number of LMIC’s providers also voiced concerns about implementation strategies to reduce PPH using oxytocin, though these were largely limited to reservations about the cold-chain storage of the drug, “….For oxytocin you have to move around with a refrigerator to keep the cold chain. In rural communities this is not possible”. (Healthcare provider Ethiopia [57]).

Line of argument synthesis.

Women who give birth and healthcare providers recognise that severe, uncontrollable blood loss in the peri- and postpartum period is life threatening. Based on their beliefs and experiences, birth attendants can identify women at risk of PPH, and most are willing to use specific skills and techniques (and, where available, treatments and uterotonic drugs) to reduce the risk. Women generally appreciate this, especially where it is done on the basis of their individual needs, beliefs and values. However, there is variation within and between healthcare providers and women, across income settings, in their views as to how far these interventions should be used routinely. In all income settings, systems for PPH prevention are likely to be more successful locally where there is a common frame of reference for the causes and consequences of PPH between service users and healthcare providers and acceptable treatments and techniques in line with this frame of reference are available free at the point of use. These systems should be applied sensitively and competently by care providers who are skilled and trained in their appropriate use; full information is available to childbearing women about side-effects of drugs and treatments and their alleviation; and where access to treatments and uterotonic drugs is not likely to result in their use in non-indicated circumstances.