Primary postpartum haemorrhage and longer-term physical, psychological, and psychosocial health outcomes for women and their partners: a mixed-methods systematic review

14

15 Most research about outcomes following postpartum haemorrhage (PPH) has focused on 16 immediate outcomes. There are fewer studies investigating longer-term maternal morbidity 17 following PPH, resulting in a significant knowledge gap. This review aimed to synthesize the 18 evidence about the longer-term physical and psychological consequences of primary PPH for 19 women and their partners from high income settings. 20

21 The review was registered with PROSPERO and five electronic databases were searched. 22 Studies were independently screened against the eligibility criteria by two reviewers and data 23 were extracted from both quantitative and qualitative studies that reported non-immediate health 24 outcomes of primary PPH. 25

26 Data were included from 24 studies, of which 16 were quantitative, five were qualitative and 27 three used mixed-methods. The included studies were of mixed methodological quality. Of the 28 nine studies reporting outcomes beyond five years after birth, only two quantitative studies and 29 one qualitative study had a follow-up period longer than ten years. Seven studies reported 30 outcomes or experiences for partners. The evidence indicated that women with PPH were more 31 likely to have persistent physical and psychological health problems after birth compared with 32 women who did not have a PPH. These problems, including PTSD symptoms and cardiovascular 33 disease, may be severe and extend for many years after birth and were more pronounced after a 34 severe PPH, as indicated by a blood transfusion or hysterectomy. There was limited evidence 35 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 3 about outcomes for partners after PPH, but conflicting evidence of association between PTSD 36 and PPH among partners who witnessed PPH. 37

38 This review explored existing evidence about longer-term physical and psychological health 39 outcomes among women who had a primary PPH in high income countries, and their partners. 40 While the evidence about health outcomes beyond five years after PPH is limited, our findings 41 indicate that women can experience long lasting negative impacts after primary PPH, including 42 PTSD symptoms and cardiovascular disease, extending for many years after birth. 43 PROSPERO registration number: CRD42020161144 44 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 4

45 Postpartum haemorrhage (PPH) is defined by the World Health Organization (WHO) as blood 46 loss of 500 ml or more from the genital tract within 24 hours after birth, (1) although the 47 definition and the cut off point for the severity of PPH varies. (2) The incidence of PPH is 48 increasing across the high-income countries. (3, 4) Most research about outcomes following PPH 49 has focused on immediate outcomes including, for example, acute organ failure, (5) blood loss, 50 hypovolemic shock and maternal death. (6, 7) There are fewer studies investigating longer-term 51 maternal morbidity following PPH, resulting in a significant knowledge gap around the 52 associations between PPH and subsequent health and wellbeing, including the impact on the 53 psychosocial and emotional wellbeing of women and their partners. (5, 8-13). 54 A systematic review conducted in 2016 investigated the prevalence of women’s emotional and 55 physical health problems following PPH and included six quantitative studies. (14) Several 56 articles investigating potential physical or psychological consequences of PPH have been 57 published since the last systematic search in April 2015. (15-17) A further systematic review 58 found a potential association between PPH and post-traumatic stress disorder (PTSD), but this 59

was limited by the small number of studies investigating this association. (18) Neither 60 of these reviews included qualitative research exploring women’s experience of PPH, suggesting 61 a comprehensive systematic review using an integrated mixed-methods approach would be 62 valuable. (19) A mixed-methods systematic review was therefore conducted to synthesize the 63 evidence about the longer-term physical and psychological consequences of PPH for women and 64 their partners. 65 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 5

66 The protocol for this review was registered with PROSPERO (CRD42020161144) and the 67 review is reported according to the Preferred Reporting Items for Systematic Review and Meta-68 Analysis (PRISMA) 2020 checklist (S1 Checklist). (20) 69 Eligibility criteria 70 We searched published literature for quantitative cohort, case-control, cross-sectional, or case 71 series studies; qualitative studies including analysis of interviews, free text boxes/open-ended 72 survey questions which explored women’s and partners’ experiences of care or seeking care and 73 information following PPH, and/or their perceptions about the physical, psychological and 74 psychosocial impact of PPH; and mixed-methods studies with an eligible qualitative or 75 quantitative component. The inclusion/exclusion criteria in Table 1 were applied. 76 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 6 Table 1. Inclusion and exclusion criteria 77 Category Inclusion criteria Exclusion criteria Population Women who gave birth after at least 24 weeks’gestation and had a PPH, and/or their partners. Women who had a PPH and/or their partners who were a subset of study participants (e.g. women who had a PPH as a subset of women with a range of maternal morbidities) unless separate data from women who had a PPH could be retrieved. Exposure For quantitative studies, primary PPH, defined as: Quantitative studies where PPH was not the primary exposure of interest or in which PPH was not well-defined and measured by one of the criteria described. Bleeding from the genital tract, within 24 hours of giving birth, with an estimated blood loss of >500 ml and/or blood transfusion within 24 hours of giving birth and/or having emergency peripartum hysterectomy (EPH) or arterial embolization following PPH, and/or peripartum fall in haemoglobin concentration of > 2g/dl. For qualitative studies, PPH regardless of definition. Comparison Studies with or without a control or comparison group, comprising participants who did not experience a PPH. None Outcomes For quantitative studies, any non-immediate and non-fatal physical, psychological and psychosocial health outcomes for women and their partners after PPH. For quantitative studies: immediate outcomes, including maternal death following PPH, acute organ failure, sepsis, shock, and with intervention outcomes such as hysterectomy, intensive care admission, blood transfusion during initial birth admission; pregnancy-related outcomes, including secondary PPH, or recurrence of PPH in subsequent pregnancies. For qualitative studies, women’s and/or partners’ experiences of care and of seeking care around the time of PPH and during the postpartum period, and their perceptions about the physical, psychological and psychosocial impact of PPH. Time For quantitative studies, any follow-up duration after hospital discharge following PPH. None For qualitative studies, all studies regardless of follow-up duration. Setting High-income countries as defined by the World Bank (21) at the time of study. Low/middle-income countries. Language English language Papers which were not written in English . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 7 Search strategy 78 Searches (S1 File) were conducted in Medline (Ovid), Embase, Web of Science, CINAHL and 79 PsycINFO on 16th December 2019 and updated on 6th January 2021 and on 31st May 2022, from 80 the inception of each database to the search date. Reference lists from all included full text 81 articles were hand searched. 82 Screening and study selection 83 Search results were imported into the reference management software Endnote and duplicates 84 were removed. (22) Search results were then imported into Covidence software for screening and 85 data extraction. (23) Titles and abstracts of all articles identified were independently screened by 86 SL and MS against the inclusion criteria. Full text screening was conducted independently by SL 87 and ME using the above eligibility criteria. Any disagreements were resolved by consensus 88 between reviewers, involving RR when necessary. 89 Data extraction 90 Data extraction was performed independently by SL, and either MS or ME. Study characteristics 91 and outcome information were extracted for both quantitative and qualitative studies using 92 bespoke data extraction forms developed and pilot-tested for this review (S2 File). For 93 qualitative studies, all the text and themes reported under findings or results from included 94 studies were imported into NVivo software and treated as a primary textual data source for 95 analysis. Data from mixed method studies were entered in both forms as appropriate. 96 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 8 Risk of bias and quality assessment 97 Risk of bias and quality assessment for the included studies was conducted by SM, MA and ME, 98 with two independent reviewers for each study, with any disagreements resolved by consensus as 99 described above. Included quantitative studies (and the quantitative component of mixed-100

studies) were assessed using the risk of bias assessment for non-randomized studies 101 (ROBANS) tool (S1 Table). (24) 102 Quality assessment for qualitative studies was conducted using the Critical Appraisal Skills 103 Programme (CASP) tool. (25) The response options for Questions 1-9 were modified from 104 yes/no/unclear to include the option to respond ‘partly’ where a paper addressed some items in 105 the checklist associated with the question, but not all. Quantitative and qualitative components of 106 mixed-methods studies were assessed and reported separately. 107 Data synthesis 108 Since the quantitative studies were not homogeneous in terms of study design and outcome 109 measurements, we used a narrative synthesis approach, presenting quantitative study results in 110 structured tables. (26) Qualitative studies were synthesised using the approach described by 111 Thomas and Harden. (27-29) This involved: 1) free line-by-line coding of the findings of 112 primary studies; 2) the organization of codes in related areas to construct ‘descriptive’ themes; 113 and 3) the development of ‘analytical’ themes to answer the research questions. (27-29) 114 We used the Joanna Briggs Institute (JBI) convergent segregated approach for mixed-methods 115 systematic reviews, whereby following separate analysis of the quantitative and qualitative 116 evidence, the similarities and differences in findings were compared to produce a narrative 117 summary. (30) 118 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 9

119 The search yielded 15,742 references, of which 7,625 were duplicates and were removed (Fig 1). 120 After removing duplicates, a total of 8,117 references remained. Following title and abstract 121 screening, 127 references were included for full text review. An additional six references were 122 identified through other sources. After removal of ineligible studies, 33 papers reporting 24 123 studies were included in the review, of which 16 were quantitative, five qualitative, and three 124 mixed-methods studies. 125 Fig 1.PRISMA flow diagram for study selection (20) 126 Characteristics of included studies 127 Included studies were conducted in Australia, New Zealand, France, the Netherlands, Sweden, 128 Switzerland, South Korea, Canada, the UK and the United States, between 2005 and 2021. The 129 follow-up time was from hospital discharge following PPH to up to six weeks postpartum in 130 seven studies, (31-38) up to a year in four studies, (12, 16, 39-42) up to five years in three 131 studies, (43-46) and longer than five years in nine studies. (47-58) One qualitative study did not 132 report on follow-up duration. (59) The definitions for PPH and severe PPH used in these studies 133 varied widely. In studies defining PPH according to blood loss volume, this varied from 500ml 134 to 1000ml for vaginal birth and from 750 ml to 1500 ml for Caesarean birth. In four studies, PPH 135 was identified by using ICD 10 codes recorded in hospital or insurance databases. (35, 42, 51, 136 52) Definitions for severe PPH also varied widely, from >1000ml to >2000ml. In seven studies 137 where the exposure was severe PPH, severity was also indicated by combining blood loss 138 volume with interventions such as emergency peripartum hysterectomy (EPH) and/or arterial 139 embolization following PPH. (39, 44, 47-49, 54, 56) PPH was not well defined in one qualitative 140 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 10 study, (38) however, all qualitative studies were included regardless of their definitions as these 141 were not part of the eligibility criteria for qualitative studies. 142 Summary characteristics of all included studies are presented in Table 2.143 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 11 Table 2. Characteristics of included studies 144 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Quantitative studies Cho 2021 (51) South Korea Merged databases of the South Korean insurance claim and health screening programme To determine whether PPH is associated with cardiovascular disease (CVD) beyond the peripartum period. Women who gave birth in 2007 Retrospective cohort Newly diagnosed CVD and/or ischemic heart disease 2007-2015 N= 150,381 Population-based study Chauleur 2008 (8) France University of Nimes Hospital To explore the effect of severe PPH and its related treatments on postpartum venous thromboembolic risk Women in their first intended pregnancy Prospective cohort Superficial venous thrombosis 1999-2004 N=32,463 Population-based study De la Cruz 2016 (44) United States Online community with 1.7 million registered users To explore if women who had emergency peripartum hysterectomy (EPH) are positive for post-traumatic stress disorder (PTSD) compared with women who did not have EPH more likely to screen Women from online EPH group Retrospective cohort PTSD, postpartum depression (PPD), both PTSD and PPD Year of study not reported N= 409 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 12 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Chessman 2018 (32) Australia New South Wales public hospital To determine the association between red blood cell transfusion and breastfeeding among women who had a PPH at birth taking into account postpartum haemoglobin concentrations Women who had a PPH and had a red blood cell transfusion following a singleton birth of at least 37weeks’ gestation. Retrospective cohort Any breastfeeding at hospital discharge Records from 2007-2010 N= 15,451 Population-based study Drayton 2016 (33) Australia All New South Wales hospitals To determine the association between red blood cell transfusion and breastfeeding at discharge among women who had a PPH during their birth admission All women giving birth in New South Wales, Australia Retrospective cohort Exclusive breastfeeding, partial breastfeeding, any breastfeeding Records from 2007-2012 N= 40,149 Population-based study Eckerdal 2016 (34) Sweden Uppsala University hospital To explore the association between PPH and PPD Women who gave birth at University Hospital in Uppsala Prospective cohort PPD 2006-2012 N=446 Population-based study . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 13 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Eggel 2021; Bernasconi 2021 (54, 60) Switzerland University Maternity Hospital in Lausanne To evaluate long term psychological, gynaecological and sexual outcomes among patients with uterine artery embolization following PPH Women who had PPH and had uterine artery embolization at University Maternity Hospital Case-control PPD, PTSD, Symptoms of menstrual disturbances, sexual dysfunction Records from 2003-2013 N=252 women, 142 partners Hospital-based study Feinberg 2005 (45) United States A tertiary hospital in Chicago To determine the incidence of Sheehan's syndrome among patients with obstetric haemorrhage Women who gave birth at North-western Memorial Hospital with obstetric haemorrhage Retrospective cohort Symptoms of Sheehan's syndrome Records from 1998-2002 N=109 Hospital-based study . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 14 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Knight 2016 (39) UK (England) 111 UK hospitals with consultant-led maternity units To explore the longer-term health outcomes of women following EPH to control haemorrhage and to see how these differ from longer-term health outcomes in women who gave birth without EPH Women who gave birth and had EPH in UK hospitals Case-control Pain, depression, anxiety, difficulty in sexual intercourse, severe tiredness, menopausal symptoms, flashbacks, difficulty in concentration 2013 using records from 2005-2006 N=78 Population-based study Liu 2021 (42) Sweden Record-linkage using Swedish national birth register linked with national patient register and prescribed drug register To examine the association between PPH and PPD Women who had a live birth between 37 weeks and 42 weeks gestational age Retrospective cohort PPD Records from 2007-2014 N=486,722 Population-based study . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 15 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Michelet 2015 (47) France Lariboisiere Hospital in France To investigate the psychological impact of EPH Women admitted with PPH who underwent EPH. Retrospective cohort PTSD, anxiety, PPD 2012 using records from 2004-2011 N=869 Hospital-based study Parry-Smith 2021a; Parry-Smith 2021b (52, 53) UK (England) NHS hospitals in England To examine long term risk of developing hypertension, CVD and mental ill health among women who had a PPH compared with women who did not Women aged between 16-46 years who gave births in England Retrospective cohort Hypertension, CVD, PPD, depression, PTSD, anxiety Records from 1990-2018 N=42,327 Population-based study Ricbourg 2015 (16) France Lariboisiere Hospital in France To investigate the psychological impact of PPH on women and their partners, including its impacts on PTSD, PPD, and the mother/child relationship Women admitted with PPH Prospective cohort PTSD, PPD, mother and infant bonding 2010-2011 N=40 women, 26 partners Hospital-based study . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 16 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Sentilhes 2011 (48) France University affiliated tertiary referral centre at Rouen University Hospital in France To estimate the longer-term psychological impact of severe PPH Women with PPH who underwent pelvic arterial embolization at the tertiary obstetric centre Retrospective cohort Psychological symptoms including negative memories 2009 using records from 1994-2007 N=68 Hospital- based study Thompson 2010; Australia and New Zealand 17 hospitals across Australia and New Zealand To explore the physical and psychological health outcomes and breastfeeding outcomes for women following PPH Women who had primary PPH in Australia and New Zealand Prospective cohort Range of physical and psychological outcomes Thompson 2011 (12, 40) 2006-2007 N=206 Population-based study . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 17 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Thurn 2018 (35) Sweden Tertiary hospitals in Sweden To investigate postpartum blood transfusion and PPH as potential risk factors for postpartum venous thromboembolism (VTE) Women who gave birth in Stockholm Retrospective cohort VTE 2016 using records from 1999-2002 N=82,376 Population-based study Ukah 2020 (50) Canada Registry containing all hospital discharges in Quebec, Canada To investigate the association between obstetric haemorrhage and CVD up to three decades after birth Women who gave birth in Quebec, Canada Retrospective cohort Hospitalisation for CVD Records from 1989-2016 N=1,224,975 Population-based study Van Steijn 2019; Netherlands 8 hospitals in Amsterdam, Netherlands To evaluate whether (i) severe PPH is a risk factor for PTSD in women and (ii) witnessing severe PPH is a risk factor for developing PTSD in partners Women with ≥ 2000 ml of blood loss and their partners in Netherlands Prospective cohort PTSD Van Steijn 2020 (36, 37) 2015-2017 N= 308 women, 185 partners Population-based study . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 18 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Van Stralen 2018 (49) Netherlands 98 hospitals with a maternity unit in the Netherlands To investigate quality of life and psychological impact of major obstetric haemorrhage on women and their partners Women who experienced major obstetric haemorrhage and had hysterectomy and arterial embolization in the Netherlands Retrospective cohort General health status and psychological outcomes 2012-2013 N=58 women, 49 partners Population-based study Qualitative studies Briley, 2020 (38) United Kingdom Postnatal wards in two South London NHS Trusts To investigate the experience of PPH, for women, birth partners, and health care professionals Women who had experienced PPH and their partners Qualitative in-depth- interview study Knowledge specific to PPH, effective and appropriate responses to PPH, communication of risk factors, and quantifying blood loss Maximum variation purposive sampling, semi-structured face to face interviews, thematic analysis Not reported N=9 women, 4 partners . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 19 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) De La Cruz 2013 (43) United States Online/internet group To explore women's peripartum experiences of EPH to make recommendations for care EPH survivors from an English- speaking international internet support group Qualitative interview study using grounded theory Death/dying, pain, fear, bonding, numbness/ delayed emotional response, communication Purposive sampling, semi-structured telephone interviews, constant comparative thematic analysis Not reported N=15 Dunning 2016 (46) United Kingdom One teaching hospital in London To investigate the experiences of women who have had a primary PPH and the experiences of birth partners who witnessed the PPH Women who had an estimated blood loss of ≥ 500 ml within 24 hours after vaginal birth and their partners who were present at the time of the PPH Qualitative interview study Control, communication, consequence (physical and psychological impact), competence Maximum variation sampling, explanatory, semi- structured face to face interviews, thematic analysis 2013-2014 N=11 women, 6 partners . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 20 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Elmir 2012a; Elmir 2012b; Elmir 2012c; Elmir 2014 (55-58) Australia University campuses and public places such as pharmacies and child care centres To describe women's experiences of having an EPH following a severe PPH including impacts on early mothering experiences and to explore the way that women find meaning and positivity in their lives Women who had an EPH following a severe PPH. Qualitative interview study using naturalistic inquiry approach Between life and death, loss of normality, early mothering experiences, moving forward 2009 N=21 Purposive snowball sampling, semi- structured face to face/telephone/e- mail/internet interviews, thematic analysis using inductive method Knight 2016 (39) United Kingdom 111 UK hospitals with maternity units To explore the longer-term health outcomes of women following an EPH to control haemorrhage Women who had an EPH to control haemorrhage Mixed method study using open-ended survey questions The staff providing care, organization / structure of care, longer-term impact of birth experiences 2013 N=78 Purposive sampling, free text questions at the end of follow-up questionnaire, thematic analysis . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 21 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Snowdon 2012 (61) United Kingdom Two hospitals in UK To explore women’s and partners’ experiences of severe PPH and its management Women who had experienced severe PPH and their partners Qualitative interview study using interpretative phenomenological approach Confidence vs fear, trust vs mistrust, satisfaction/ dissatisfaction, communication difficulties 2006-2007 N=9 women, 6 partners Purposive sampling, semi-structured face to face interviews, thematic analysis Thompson, 2010; Thompson, 2011 (12, 41) Australia and New Zealand 17 hospitals with maternity services in Australia and New Zealand To identify sources of distress for women and gaps in service provision, particularly their informational needs Women who had severe PPH Mixed-method study using open-ended survey questions Adequacy of care, emotional responses to the experience, implications for the future, and concerns for their baby 2006-2007 N=206 Purposive sampling, open-ended self- completed survey questions, thematic analysis . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 22 Study Country Setting Aims Population Study design Outcome(s) / Year of study Sample size Themes reported (for qualitative studies) Woiski 2015 (59) Netherlands, 3 University Hospitals in Netherlands, and childbirth forums on internet To identify obstacles and facilitators for providing high quality PPH care, from both patient and professional perspectives Women who gave birth and lost more than 1000 ml of blood after birth Qualitative interview study Influencing factors from patient perspectives, professional factors and organizational factors Not reported N=12 Purposive maximum variation sampling, semi-structured face to face/ telephone interviews with patients, framework analysis . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 23 Quantitative studies 145 Risk of bias assessment 146 The results of risk of bias assessment for quantitative studies and quantitative components of 147 mixed-methods studies are presented in Table 3. Most included studies (n=15) had a low risk of 148 bias (i.e., low risk in most domains assessed). However, four studies had a high risk of bias in 149 two or more areas of domains assessed, with the main sources of potential bias being participant 150 selection, outcome measurement and incomplete data. (44, 47, 49, 54) 151 Table 3 Risk of bias for quantitative studies using ROBANS (24) 152 Study Participant selection Exposure measurement Confounding Outcome measurement Incomplete data Se rep Chauleur 2008 (8) Chessman 2018 (32) Cho 2021 (51) De la Cruz 2016 (44) Drayton 2016 (33) Eckerdal 2016 (34) Eggel 2021; Bernasconi 2021 (54, 60) Feinberg 2005 (45) Knight 2016 (39) Liu 2021 (42) Michelet 2015 (47) 23 Selective reporting . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 24 Parry-Smith 2021a; Parry-Smith 2021b (52, 53) Ricbourg 2015 (16) Sentilhes 2011 (48) Thompson 2010; Thompson 2011(12, 40) Thurn 2018 (35) Ukah 2020 (50) Van Steijn 2019; Van Steijn 2020 (36, 37) Van Stralen 2018 (49) Low risk of bias , Unclear, High risk of bias

of quantitative synthesis 153 Quantitative findings are summarised separately for physical and psychological health outcomes 154 for women and their partners in Table 4. 155 24 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 25 Table 4. Physical and psychological health outcomes investigated by included quantitative studies 156 Study Sample size (N) Time since PPH Outcome Outcome assessment Percentage or median/mean (IQR/SD) Percentage or median/mean (IQR/SD) Risk ratio (RR) or Odds ratio (OR) or Hazard ratio (HR) PPH No PPH (95% CI; p value) Physical health outcomes Thompson 2011 (40) 206 2 months General Health Median SF36 75 (60–80) NC NC 4 months 72.5 (60–80) NC NC Van Stralen 2018 (49) 58 6-9 years Mean SF36 70 (23.7) 71.5 NR (p=0.23) SF36 reference group Drayton 2016 (33) 40,149 Hospital Discharge Exclusive breastfeeding Perinatal database 81% (PPH only) 83% NR 71% (PPH with transfusion) Chessman 2018 (32) 15,451 Hospital discharge Perinatal database 67% 79% NR (Hb >90g/dl) 71% (Hb 70-90g/dl) 65% (Hb<70 g/dl) Knight 2016 (39) 78 7 days Health questionnaire 39% 60% NR (p=0.089) 1.5 months 14% 40% NR (p=0.018) 6 months 12% 19% NR (p=0.519) Thompson 2010 (12) 206 7 days Health questionnaire 63% NC NC 2 months 58% NC NC 4 months 45% NC NC Chauleur 2008 (8) 32,463 1.5 months VTE Colour duplex ultrasound NR NR RR=5.4 (1.68, 17; p=0.0047) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 26 Study Sample size (N) Time since PPH Outcome Outcome assessment Percentage or median/mean (IQR/SD) Percentage or median/mean (IQR/SD) Risk ratio (RR) or Odds ratio (OR) or Hazard ratio (HR) PPH No PPH (95% CI; p value) aRR=5.3 (1.6, 17; p=0.005) Thurn 2018 (35) 82,376 1.5 months ICD 10 diagnostic codes NR NR OR=2.85 (1.47, 1.55; p=0.001) aOR=2.7 (1.4, 5.2) aOR=1.4 (0.5, 3.5) (Further adjusted for blood transfusion) Cho 2021 (51) 150,381 Up to 8 years CVD ICD 10 diagnostic codes 4.5% 4.6% aHR=0.96 (0.86, 1.07) (PPH) (PPH) 6.9% aHR=1.60 (1.25, 2.06) (PPH with transfusion) (PPH with transfusion) Ukah 2020 (50) 1,224,975 Up to 29 years ICD 10 diagnostic codes NR NR aHR=1.00 (0.95, 1.06) (PPH) aHR=1.38 (1.13, 1.68) (PPH with transfusion) Parry-Smith 2021a (52) 42,327 Up to 8 years Diagnostic codes from English primary care database 0.16% 0.18% aHR=0.86 (0.52, 1.43; p=0.57) Feinberg 2005 (45) 109 12-40 months Sheehan's Syndrome Two or more symptoms in questionnaire 25% 7% NR Knight 2016 (39) 78 12 months Difficulty in sexual intercourse Health questionnaire 27% 6% NR 8.1 years 13% 8% NR . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 27 Study Sample size (N) Time since PPH Outcome Outcome assessment Percentage or median/mean (IQR/SD) Percentage or median/mean (IQR/SD) Risk ratio (RR) or Odds ratio (OR) or Hazard ratio (HR) PPH No PPH (95% CI; p value) Eggel 2021 (54) 252 8.1 years Sexual dysfunction Mean FSFI score 23.8 (0.4) 23.2 (0.6) NR Knight 2016 (39) 78 12 months Menopausal symptoms Health questionnaire 17% 0% NR 8.1 years 33% 13% NR Knight 2016 (39) 78 12 months Pain Health questionnaire 43% 19% NR 8.1 years 10% 10% NR Knight 2016 (39) 78 12 months Severe tiredness and Fatigue Health questionnaire 33% 13% NR 8.1 years 17% 21% NR Thompson 2011 (40) 206 2 months Postpartum fatigue Median Milligan score 17 (15-18) NC NC 4 months 15 (14-17) NC NC Thompson 2011 (40) 206 2 months Mastitis Health questionnaire 22% NC NC 4 months 14% NC NC Van Stralen 2018 (49) 58 6-9 years Patient Perception- Consequences Mean B-IPQ 7.5 (2.6) NC NC Eggel 2021 (54) 252 8.1 years Amenorrhoea Health questionnaire 12% 14% NR Dysmenorrhoea 22% 25% NR Spotting 7% 7% NR Psychological health outcomes Michelet 2015 (47) 869 26.5 months Anxiety Median HADS 10 (8-12) NC NC Thompson 2011 (40) 206 2 months Median STAI scale (State subscale) 10 (9-11) NC NC 4 months 10 (9-11) NC NC . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 28 Study Sample size (N) Time since PPH Outcome Outcome assessment Percentage or median/mean (IQR/SD) Percentage or median/mean (IQR/SD) Risk ratio (RR) or Odds ratio (OR) or Hazard ratio (HR) PPH No PPH (95% CI; p value) Knight 2016 (39) 78 12 months Health and wellbeing questionnaire 27% 10% NR 8.1 years 17% 17% NR Parry-Smith 2021b (53) 42,327 Up to 8 years Diagnostic codes from English primary care database 5% 5% aHR=0.99 (0.90, 1.09; p=0.881) Michelet 2015 (47) 869 26.5months Postpartum depression / Depression Median HADS scale 6 (2-8) NC NC Knight 2016 (39) 78 12 months Health and wellbeing questionnaire 33% 8% NR 8.1 years 13% 19% NR Delacruz 2016 (44) 409 6 months EPDS (score ≥ 12) 6% 9% NR 36 months 0% 4% NR Eckerdal 2016 (34) 446 1.5 months EPDS (score ≥ 12) 14% 10% aOR=1.81 (0.91, 3.57) Ricbourg 2015 (16) 40 1 months EPDS (score ≥ 12) 35% 15% NR 3 months 11% 0% NR Thompson 2011 (40) 206 2 months EPDS (score ≥ 12) 11% NC NC 4 months 13% NC NC Liu 2021 (42) 486,722 12 months ICD-10 codes for depression or PPD, or prescription record of antidepressants NR NR aHR=1.08 (0.99, 1.17) Parry-Smith 2021b (53) 42,327 Up to 8 years Diagnostic codes from English primary care database 5.34% 4.75% aHR=1.10 (1.01, 1.21; p=0.037) Bernasconi 2021 (60) 142 8 years MINI score 4.1 (0.5) 2.7 (0.3) NR, (p=0.015) De la cruz 2016 (44) 409 6 months Post-traumatic stress disorder PSS score ( ≥ 15) 9% 1% aOR=2.46 (1.92, 3.16) 36 months PSS score (≥ 15) 24% 5% aOR=1.90 (1.57, 2.30) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 29 Study Sample size (N) Time since PPH Outcome Outcome assessment Percentage or median/mean (IQR/SD) Percentage or median/mean (IQR/SD) Risk ratio (RR) or Odds ratio (OR) or Hazard ratio (HR) PPH No PPH (95% CI; p value) Michelet 2015 (47) 869 26.5 months IESR (≥ 30) 64% NC NC Ricbourg 2015 (16) 40 1 months IESR (≥ 30) 45% 21% NR 3 months 24% 13% NR Thompson 2011 (40) 206 2 months PCL (>44) 5% NC NC 4 months 3% NC NC Van Steijn 2020 (37) 308 1-1.5 months Post-traumatic stress disorder PCL-5 ≥ 32) 7% 1.7% RR=4.45 (0.99, 20.06; p=0.035) 1.5 months CAPS-5 5.60% 0% NR, (p=0.007) Parry-Smith 2021b (53) 42,327 12 months Diagnostic codes from English primary care database NR NR aHR=3.44 (1.31, 9.03) Up to 8 years 0.20% 0.17% aHR=1.17 (0.73, 1.89; p=0.511) Bernasconi 2021 (60) 142 8 years TSQ (≥ 7) 3.5 (0.4) 2.0 (0.2) aOR=5.1 22.2% 4.8% (1.5, 17.5; p=0.001) Ricbourg 2015 (16) 40 1 months Mother-infant bonding Median Mother Infant Bonding Scale (MIBS) 1 (0-2) 0 (0-2) NR 3 months 1 (0-1) 0 (0-2) NR Thompson 2011 (40) 206 2 months General mental health Median SF36 80 (70–80) NC NC 4 months 80 (70–90) NC NC Van Stralen 2018 (49) 58 6 -9 years Mean SF36 75 (17.4) 75.5 NR (p=0.85) Knight 2016 (39) 78 12 months Difficulty in concentrating Health questionnaire 23% 2% NR 8.1 years 23% 17% NR Knight 2016 (39) 12 months Flashbacks to labour or birth Health questionnaire 33% 6% NR 8.1 years 17% 2% NR Sentilhes 2011 (48) 68 1-15 years Negative memories of PPH Survey question 68% NC NC . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 30 Study Sample size (N) Time since PPH Outcome Outcome assessment Percentage or median/mean (IQR/SD) Percentage or median/mean (IQR/SD) Risk ratio (RR) or Odds ratio (OR) or Hazard ratio (HR) PPH No PPH (95% CI; p value) Sentilhes 2011 (48) 1-15 years Suggestion by hospital staff of a visit to psychologist Survey question 66% NC NC aRR= adjusted risk ratio, aOR= adjusted odds ratio, aHR=adjusted hazard ratio NC= No comparison group, NR= Not reported Outcome measurements: SF 36 (Short form health survey): Scores 0-100 with a higher score reflecting better outcomes Milligan score: Scores 10–14 ‘no⁄low’ postpartum fatigue, 15–20 ‘medium’, and 21–40 ‘high’ postpartum fatigue B-IPQ (Brief Illness Perception Questionnaire): Scores 0-80 with higher scores reflecting a greater perceived burden of a disease and lower scores reflecting better outcomes FSFI (Female Sexual Function Index Questionnaire): Scores 2-36 with higher scores indicating high sexual activity and satisfaction, and lower scores indicating sexual difficulties or low sexual activity HADS (Hospital Anxiety and Depression Scale): Scores 0-21 with 0-7 normal, 8-10 borderline abnormal, 11-21- abnormal STAI (Spielberger State-Trait Anxiety Inventory): Scores 6-24 with >12 indicating high anxiety EPDS (Edinburgh Postnatal Depression Scale): Scores 0-30 with >12 indicating postnatal depression. MINI (Mini International Neuropsychiatric Interview): Scores 0-10 PSS Scale (Post-traumatic stress disorder symptom scale): 17 item checklist with ≥ 15 indicating PTSD IESR Scale (Impact of Events Scale): Scores 0-88 with ≥ 30 indicating PTSD PCL scale (17 item checklist): Scores 17-85 with >44 indicating PTSD PCL-5 score (20 item self-report tool): Scores 0-80 with ≥ 32 indicating PTSD CAPS-5 score (Gold standard to diagnose PTSD using criterion A, B-E, F, G): Having at least one B Criterion symptom, one C Criterion symptom, two D Criterion symptoms and two E Criterion symptoms indicates PTSD. TSQ (Trauma Screening Questionnaire): Score 0-10 with ≥ 7 indicating PTSD Mother Infant Bonding Scale: Scores 0-27 with a high score indicating a disorder related to mother-child bonding. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 31 Physical health outcomes for women 157 Two studies reported on general health status in women following PPH. (40, 49) While the first 158 found a decline in general health status at two months and four months following the PPH, (40) 159 the second found that there was no difference in SF36 score for women following PPH compared 160 to the SF36 reference group values. (49) Neither study collected data from an appropriate 161 comparison group. 162 All four studies that explored breastfeeding as an outcome reported that there was no significant 163 difference in the proportion of exclusive breastfeeding at hospital discharge among mothers who 164 had a PPH, compared with mothers who did not have a PPH. (12, 32, 33, 39) However, the 165 proportion of exclusive breastfeeding at hospital discharge (and at six weeks and six months after 166 birth) was slightly lower among mothers who received a blood transfusion, had a low 167 Haemoglobin level or an EPH following PPH, compared with women who did not have a PPH in 168 three studies. (32, 33, 39) One study did not have a comparison group of women who did not 169 have a PPH. (12) 170 Two studies explored the potential association between PPH and venous thromboembolism 171 (VTE) at six weeks postpartum. (8, 35) In one study, severe PPH was strongly associated with 172 the occurrence of VTE in the first six weeks after birth after adjusting for caesarean section and 173 low molecular weight heparin use after birth (OR 5.3, 95% CI: 1.6-17, p=0.005). (8) In the 174 second, the authors concluded that PPH was not a major risk factor for VTE unless there was an 175 associated blood transfusion following PPH. (35) 176 Three studies explored the potential association between PPH and the risk of CVD up to 29 years 177 after birth. (50-52) Two studies suggested that women with PPH requiring blood transfusion had 178 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 32 a higher risk of hospitalization for CVD (aHR 1.60, 95% CI 1.25-2.06 and aHR 1.38, 95% CI 179 1.13, 1.68) compared with women without PPH, after adjusting for potential confounders. (50, 180 51) However, the third study did not find an association between PPH and hypertension or CVD. 181 (aHR 1.03, 95% CI 0.87-1.22) (52) 182 Other physical outcomes investigated, each addressed by one study, included Sheehan’s 183 syndrome (damage to the pituitary gland following major haemorrhage), (45) difficulty in sexual 184 intercourse, (39) menopausal symptoms, (39) menstrual disturbances, (54) pain, severe tiredness 185 (39) and fatigue, and mastitis. (12) A higher proportion of women reported these symptoms in 186 severe PPH or EPH groups compared to women who did not have a PPH or EPH. 187 Psychological health outcomes for women 188 Four studies reported the prevalence of anxiety following PPH at different time points. (39, 40, 189 47, 53). The exposed group in these studies was women who had a PPH or severe PPH and the 190 follow-up time ranged from two months to eight years after birth. The score for the Hospital 191 Anxiety and Depression Scale (HADS) in the first study indicated moderate anxiety among 192 women who had EPH at median follow-up time of 26.5 months after birth. (47) The second 193 study showed there was mild anxiety among women who had a PPH at both two month and four 194 month follow-up times. (40) Both of these studies did not have comparison groups. The results 195 from the third study indicated women who had hysterectomy were more likely to report 196 anxiety/nerves (27%) compared with women who did not have a PPH (10%) in the first 12 197 months after birth. Another study, with a follow-up duration of up to 8 years, did not find any 198 association between PPH and anxiety. (53) 199 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 33 Nine studies explored potential associations between PPH and postpartum depression (PPD) or 200 depression. (16, 34, 39, 40, 42, 44, 47, 53, 60) with the follow-up time ranging from one month 201 to eight years after birth. Taken together, the findings of these studies were inconclusive, 202 suggesting different directions of association between PPH and depression. Two studies reported 203 a slightly higher prevalence of depression in the PPH group compared with the non-PPH group 204 at one, three and 12 months after birth. (16, 34) and another two studies with a follow-up of up to 205 8 years found a higher mean depression score among women who had UAE following PPH, (60) 206 and a 10% increased risk of developing PPD among women who had a PPH compared to women 207 who did not have a PPH after adjusting for age, BMI, smoking, ethnicity, birthweight and 208 delivery method (aHR 1.10, 95% CI 1.01-1.21, p 0.037). (53) The results from multivariable 209 regression models from two further studies suggested no statistically significant association 210 between PPH and PPD at six weeks and 12 months after birth. (34, 42), while a further study 211 reported a slightly higher prevalence of depression in the non-PPH group (6% in PPH vs 9% in 212 non-PPH) at six months after birth. (44) Two studies did not have comparison groups, but one 213 reported a median HAD score of 6 (normal) while the other reported 11% and 13% of women 214 who had a PPH and/or EPH had symptoms of postpartum depression at two and four months 215 after birth. (40, 47) 216 Seven studies explored the association between PPH and post-traumatic stress disorder (PTSD) 217 at different time points ranging from 1 month to 8 years after birth, indicating a potential 218 association between PPH and PTSD. (16, 37, 40, 44, 47, 53, 60) Five studies found a higher risk 219 of PTSD among women who had a PPH and/or UAE or EPH compared with women who did not 220 experience a PPH. (16, 37, 44, 53, 60) Four of these studies also reported adjusted odds (aOR), 221 risk ratios (aRR) or aHR and these studies suggested a significant association between PPH and 222 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 34 PTSD up to eight years after birth [aRR=4.45 (1.5 month); aRR=2.46 (6 months); aHR= 3.44 223 (within 12 months); aRR=1.90 (36 months); aOR=5.1 (8 years)]. (37, 44, 53, 60) The other two 224 studies did not have comparison groups, but reported respectively 5% prevalence of PTSD 225 among women who had a PPH at two months after birth (40) and 64% prevalence of PTSD 226 among women who had an EPH at the median follow-up of 26.5 months. (47) 227 Other psychological health outcomes investigated included mother-infant bonding, (16) general 228 mental health score, (40, 49) difficulty in concentrating, flashbacks, (39) negative memories of 229 PPH, and suggestion by hospital staff of a visit to psychologist. (48) Results indicated that a 230 higher proportion of women in severe PPH or EPH groups reported having difficulties in 231 concentrations and flashbacks compared with women who did not have a PPH or an EPH. 232 Physical and psychological health outcomes for partners 233 Three quantitative studies reported on the prevalence of psychological health outcomes among 234 partners of women who had experienced PPH. (36, 49, 60) The first study reported that there was 235 no strong association between PTSD diagnosis and witnessing PPH among partners of women 236 according to a self-report questionnaire. (36) The second study explored the quality of life for 237 partners of women who had experienced severe PPH using SF-36 general health survey tools. 238 (49) The results indicated that partners of women who had severe PPH had better scores overall 239 compared with the reference group’s scores from myocardial infarct and systemic lupus 240 erythematosus patients. In contrast to these studies, the third study reported that there was a 241 higher prevalence of depression and PTSD among partners of women who had PPH compared to 242 partners of women without PPH (11.5% versus 1.5%, p=0.019). (60) 243 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 35 Qualitative studies 244 Quality assessment 245 The results of quality assessment for the eight included qualitative studies and qualitative 246 components of mixed-methods studies are presented in Table 5. All but one of the studies had a 247 response of “yes” to at least seven of the questions in the CASP tool, with the remaining study, 248 having a “partially” response for three questions. (39) Only two of eight studies (46, 61) reported 249 the relationship between researchers and participants clearly (i.e., Q6), so it was not possible to 250 assess whether the roles of researchers had a potential impact on the research study and 251 interpretation of findings. 252 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 36 Table 5. Quality assessment for qualitative studies using CASP tool (25) 253 Study 1.Clear statement of aims 2.Qualitative methodology appropriate? 3. Research design appropriate for the aims? 4.Recruitme nt strategy appropriate ? 5. Data Collection addressed the research issue? 6.Relationshi p between researcher and participants adequately considered? 7.Ethical consideration 8.Data analysis rigorous? 9. A clear statement of findings? 1 v th r Briley 2020 (38) M De La Cruz 2013 (43) M Dunning 2016 (46) H Elmir 2012a; Elmir 2012b; Elmir 2012c; Elmir 2014 (55-58) M Knight 2016 (39) M Snowdon 2012 (61) H Thompson 2010; Thompson 2011 (12, 41) H Woiski 2015 (12, 41, 59) L Items 1-9 scored as Yes , Partially , Can’t tell , No Item 10 scored as low, moderate, or high value 36 10. How valuable is the research? Moderate Moderate High Moderate Moderate High High Low . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 37

of qualitative synthesis 254 Analytical themes are presented under four headings: physical impact, psychological impact, 255 psychosocial impact, and experiences of care. Fig 2 presents the overall structure of the themes 256 arising from the synthesis, organized into four key time periods: during the emergency, 257 immediate hospital recovery, postnatal period up to six months after birth, and longer-term. 258 Fig 2. Summary of analytic themes from thematic synthesis of qualitative studies (n=8) 259 Women’s experiences 260 Physical impact 261 Painful recovery 262 This theme was contributed to by five studies and encompasses women’s experiences of pain at 263 the time of the PPH and for up to a year afterwards. (39, 41, 43, 46, 56) Physical limitations 264 during the immediate postnatal period and beyond were consistently reported by women across 265 five studies. (39, 41, 43, 46, 56) Their descriptions indicated that pain prevented them from 266 performing simple day to day activities such as walking and cooking dinner. (46, 56) One 267 particular concern raised by women was that they were not able to take proper care of their 268 babies as they were not fully recovered from pain. (41) In the longer-term, one woman who 269 experienced an EPH reported that it took a year for her to physically recover. (39) 270 Obstacles to breastfeeding 271 This theme was contributed to by three studies. (12, 43, 55) During their immediate hospital 272 recovery, women reflected that their initial separation from their baby when they were admitted 273 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 38 to intensive care was one of the key barriers to initiating breastfeeding. (12, 55) Other key 274 obstacles reported by women included pain, fatigue, limited mobility and posture problems, and 275 delayed milk production. (12, 55) Women also talked about how their efforts and willingness to 276 breastfeed after PPH were not sufficiently supported by health care professionals, including for 277 example being told that they should not be bothered to breastfeed after a traumatic childbirth (12, 278 43) 279 Psychological impact 280 Brush with death 281 This theme was contributed to by three studies and encompasses women’s acute psychological 282 distress as a result of an unexpected encounter with an emerging life-threatening event. (41, 43, 283 56) 284 For many women, being so close to death was traumatic and frightening, particularly when 285 women did not anticipate that death was a possibility. (41, 43, 56) Their, overwhelmingly 286 negative, emotional responses to this were described as varying from “shock”, to having a 287 “major anxiety attack” or having an “out of body experience”. (41, 43, 56) Elmir et al. (2012) 288 also highlighted that women were afraid of “leaving their young children behind to the 289 unknown” and not knowing “whether or not they would cope without a mother”. 290 Symptoms of post-traumatic stress and long term recovery 291 This theme was contributed to by five studies and describes women’s experiences of post-292 traumatic stress symptoms (PTSD) following PPH. (39, 41, 43, 46, 56, 62) PTSD is characterised 293 by four key symptoms: reliving experiences, avoidance of reminders of trauma, negative 294 thoughts and mood, and hyperarousal. (14) 295 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 39 Women described re-experiencing their traumatic birth in the form of vivid memories, 296 flashbacks and nightmares. (41, 56) Triggers for flashbacks included seeing other pregnant 297 women, being in a hospital environment or seeing stories in the media about giving birth. (56) 298 Women also talked about avoiding reminders of trauma. One woman, for example, said that she 299 avoided hospital environments. (56) For some women, reported in two studies, these symptoms 300 were severe, including having suicidal thoughts. (39, 56) 301 While the overwhelming majority of women’s observations from most studies were negative, 302 women from one study reflected on being able to take something positive from their experiences, 303 talking about how it helped them find some perspectives and resilience. (62) 304 Psychosocial impact 305 Interrupted bonding 306 This theme was contributed to by three studies and reflects women’s perceptions of bonding with 307 their infants and their role as a mother during their recovery in hospital and in the postnatal 308 period. (12, 43, 55) This theme is interlinked with the theme: obstacles to breastfeeding. 309 Women reported feeling that bonding with their babies was disrupted by physical separation 310 because of their admission to intensive care or their babies’ admission to neonatal care. Women 311 talked about having concerns for their babies’ wellbeing and not knowing if they were okay. (12, 312 55) Women reflected that this initial separation prevented them from holding, touching and 313 breastfeeding their babies, and many felt that this was a key barrier establishing a close 314 relationship with their babies. 315 Women’s reports in these studies showed that this initial interrupted bonding could continue in 316 the weeks following hospital discharge for mothers, some of whom experienced extreme 317 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 40 tiredness, continuing for up to a year after childbirth. Some women in one study talked about 318 feeling “upset” and “guilty” when their caring responsibilities were taken away by their family 319 members and partners. (55) 320 Loss of future family 321 This theme was contributed by two studies and encompasses women’s feelings of loss in relation 322 to future childbearing following PPH. (39, 41) At four months after childbirth, some women 323 talked about feeling worried about having a PPH in any future births and losing their confidence 324 in their ability to go through birth again. (41) Traumatic birth experiences convinced some 325 women not to have more children. (41) Women who had a hysterectomy following PPH talked 326 about their regret at not being able to have more children. (39) 327 Relationship stress with their partners 328 This theme was contributed to by three studies and reflects women’s perceptions of their 329 relationships with their partner following PPH and/or hysterectomy, in the immediate postnatal 330 period and up to a year after birth. (39, 46, 57) In one study, one woman who had a PPH 331 described having a relationship breakdown with her partner following a traumatic birth. (46) 332 Women who had a hysterectomy following PPH talked about sexual tension and intimacy issues 333 with their partners. (39, 57) For instance, a woman in one study mentioned loss of libido and 334 avoidance of physical contact with her partner, (39) with this lack of sexual desire and “fear for 335 intimacy” also being echoed by women in a second study. Women also reported developing 336 insecurities over time causing a strain in their marriage. (57) 337 338 339 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 41 Loss of female identity 340 This theme was contributed to by two studies and covers women’s experiences of hysterectomy 341 as a consequence of a PPH. (39, 57) Women’s accounts reflected that the memories lingered 342 with them for many years and this changed how they felt about themselves as a woman. (57) For 343 some women the lack of uterus was associated with feelings of incompleteness and emptiness. 344 (57) This was echoed by women in the second study as they talked about how hysterectomy 345 made them feel like a different person and not like a “complete woman”. (39) 346 Experiences of care 347 Adequacy of care 348 This theme was contributed to by six studies. Both positive and negative experiences of care 349 were reported by women, but all demonstrated the crucial role of health care professionals 350 helping women feel calm and supported in an emergency when they established mutual trust. 351 (38, 39, 41, 43, 46, 59) Some women reflected that they appreciated it when the doctors and 352 midwives appeared confident and calm, and took time to reassure them afterwards. (38, 46) 353 Others described feeling unsupported when midwives were “impatient” towards them, “too 354 busy” to explain things or did not “acknowledge” them while providing care. Women’s accounts 355 also showed how the care environment could affect their overall experiences. One woman, for 356 example, reported that she was very pleased to be in the critical care unit compared to a busy 357 ward. (39) 358 Insufficient follow-up care and limited resources for psychological health during the postnatal 359 period were identified in four studies. (38, 39, 41, 43) Women’s accounts indicated a need for 360 psychological counselling and follow-up debrief during the postnatal period. (39, 41, 43) Some 361 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 42 women also mentioned that patient information websites and support for breastfeeding would be 362 helpful for them. (41, 59) 363 Communication gaps and information needs 364 This theme was contributed to by six studies and describes women’s information and 365 communication needs and preferences, including their information seeking behaviours. (38, 39, 366 41, 43, 46, 59, 61) It is important to provide women who had experienced PPH a thorough 367 explanation and clear communication about their condition, treatments received including 368 hysterectomy and its potential consequences. (39) A lack of information about their current 369 health condition could worsen psychological distress for women and their partners. (43, 61) 370 During the emergency, women appreciated receiving a full explanation and information about 371 their hysterectomy and the indications for it before they consented to have an EPH. Women from 372 five studies consistently reported that they would have liked to know about available treatment 373 options and consequences of PPH, including recovery time after they had experienced PPH. (38, 374 41, 43, 46, 61) 375 During the immediate recovery period, women’s accounts demonstrated that most women were 376 unaware of excessive blood loss. (38, 46) They wanted information about what had happened to 377 them, their current location (e.g. ICU or postnatal ward), and the condition of their babies in 378 simple terms without using confusing technical words. (41, 43, 61) Another communication gap 379 reported by women who had a hysterectomy was that nurses were not aware of their complicated 380 birth and women found it “unsettling” when they had to explain to that they did not have a uterus 381 for examination, and had no need for contraception. (43) 382 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 43 Women’s information needs extended into the postnatal period, and included communication 383 about the implications of PPH and recovery time, impacts on breastfeeding, and available 384 support services. (41, 43, 46, 59) 385 Partners’ experiences 386 Three qualitative studies also included partners of women as participants. (38, 46, 61) Partners’ 387 accounts showed that, while women were facing life and death situations, partners also 388 experienced psychological distress from witnessing this. (46, 61) One partner described himself 389 as “desperate” at being left alone with his baby without support or communication about what 390 had happened to the baby’s mother. (61) This was also echoed by others as “distressing”, and 391 one partner reported seeking help from the GP and counselling services in the months after 392 witnessing such a traumatic birth. (46) 393 During the emergency and the immediate recovery period, partners appreciated it when doctors 394 took time to reassure them and did not find it helpful when health care staff appeared to show 395 panic or went quiet. (38) In terms of information and communication needs and preferences 396 around the time of the PPH, (38, 46, 61) partners wanted more information about available 397 treatment options and potential implications while waiting to find out what had happened. (46, 398 61) They reported feeling frustrated when they were being asked to leave the emergency room 399 without being fully informed about what was happening. (46, 61) Some suggested that an 400 information sheet about PPH would be helpful for partners and family members to be able to 401 learn more about the situation. (61) 402 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 44

403 The main aim of this review was to identify and summarise the available quantitative and 404 qualitative evidence about longer-term physical, psychological and psychosocial health outcomes 405 for women and their partners following primary PPH in high-income countries. We included 24 406 studies with a mix of methodological quality ranging from weak to strong. Follow-up time 407 ranged from hospital discharge following PPH to 29 years. Among these, nine studies explored 408 the longer-term impacts of primary PPH after five years, and just two quantitative studies (48, 409 50) and one qualitative study (55-58) had a follow-up period longer than ten years. Therefore, 410 there is limited research about longer-term health outcomes following primary PPH compared 411 with the many studies that have investigated acute morbidity and health outcomes related to 412 primary PPH. (7) 413 Our review provides evidence that women with PPH are more likely to have persistent physical 414 and psychological health problems during the postpartum period compared with women who 415 have not had a PPH. These differences were more pronounced for women who had some 416 indication of severe PPH such as a blood transfusion, or an EPH. The quantitative findings 417 suggest that women with severe PPH are more likely to have a long recovery time, low general 418 health score, symptoms of tiredness/fatigue, and are less likely to exclusively breastfeed. These 419 quantitative findings are consistent with women’s descriptions of their physical functioning and 420

in performing simple day-to-day activities and women’s experiences of challenges 421 around initiating and maintaining breastfeeding from qualitative studies. Our review’s findings 422 build upon the findings of a previous review conducted in 2016 by including updated literature 423 and by synthesizing available qualitative evidence on women’s and partners’ experiences of PPH 424 illuminating the different directions and size of effects in quantitative studies. (14) 425 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 45 A survey conducted among 372 women who had major obstetric haemorrhage in the 426 Netherlands, in which PPH was not distinguished from an antepartum haemorrhage, indicated 427 that 28% of the participants experienced longer-term negative impacts up to six years after birth. 428 (63) Major obstetric haemorrhage was found to have a negative impact on the partner and family, 429 and on their work, and 25% of these women reported having an additional absence from work in 430 addition to maternity leave. (63) 431 Bearing in mind the limitations identified in terms of the quantity and quality of the evidence, 432 our review indicates that some longer-term physical and psychological health outcomes after 433 PPH may be severe and extended several years after childbirth. Included quantitative studies 434 indicated that the prevalence of PTSD symptoms ranged from 3% to 64% among women with 435 PPH. Women’s descriptions from qualitative studies, which included key symptoms of PTSD 436 such as flashbacks and nightmares of having PPH, were supportive of these quantitative findings. 437 These PTSD symptoms were commonly present among women who had severe PPH and lasted 438 for years after birth. This finding is supported by qualitative studies which concluded that 439 childbirth related PTSD can have severe and lasting effects on women and their relationships 440 with family. (64-66) In the general population, the mean prevalence of postpartum PTSD is 4%-441 6% and may be higher, with a mean prevalence of 19% among women with physical 442 complications and mental health problems such as depression. (67, 68) While our review’s 443 findings are indicative of a higher prevalence of PTSD symptoms following primary PPH, (16, 444 37, 44, 60) more evidence is required before concluding a causal association between PPH and 445 PTSD. (18) 446 Our review included three studies which explored the association between PPH and CVD, and 447 the findings suggested that women who had a severe PPH requiring a transfusion were more 448 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 46 likely to develop CVD in later life. Prior studies have suggested that women with severe PPH are 449 at a higher risk of acute heart failure and myocardial ischaemia which results from haemorrhagic 450 shock. (69, 70) Therefore, there is a potential that this acute failure has implications for the 451 cardiovascular system which may put women at higher risk of developing subsequent CVD in 452 later life. Other obstetric complications, including pre-eclampsia, preterm birth, maternal 453 anaemia, and gestational diabetes, have also been shown to increase the subsequent risk of 454 developing CVD. (71-74) It is therefore important to consider women’s obstetric history before 455 reaching conclusions about a casual association between PPH and CVD. 456 Qualitative evidence included in this review indicated that whether women received adequate 457 support during the immediate recovery and postnatal period may affect their acute psychological 458 wellbeing and have an impact on longer-term psychological health outcomes after PPH. For 459 example, women’s accounts suggested that they received insufficient follow-up care and had 460 unmet needs for breastfeeding and psychological health services. Health care professionals and 461 family members were crucial in helping women feel calm and supported. This is supported by 462 qualitative studies conducted in Australia which have described how care providers’ interactions 463 and postnatal support from partners can affect women’s experiences of trauma during childbirth. 464 (75-77) 465 We found very limited evidence about outcomes for partners after PPH, with only three 466 quantitative studies without comparison groups. The evidence from one study suggested an 467 increased risk of PTSD among partners who witnessed PPH while another two studies did not 468 indicate any strong associations between PTSD or general health status. However, partners’ 469 accounts in qualitative studies indicated that they experienced acute psychological stress and 470 anxiety due to inadequate information when women were in the emergency room, and some of 471 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 47 them required counselling services after witnessing PPH. PPH may also have an impact on 472 relationships between women and their partners. Witnessing their partner having PPH and 473 having relationship difficulties may affect partner’s psychological wellbeing beyond the 474 postnatal period. Qualitative studies into fathers’ experiences who witnessed birth trauma have 475 identified that this can affect their mental health and relationships with friends and family long 476 into the postnatal period. (78, 79) Studies have also found that men are more likely to cope with 477 these traumatic experiences by avoidance, and may be reluctant to seek necessary care which 478 may lead to poor psychological health outcomes. (80, 81) Therefore, it is crucial to investigate 479 the potential psychological impact among partners who have witnessed a PPH. 480 Strengths and limitations 481 To our knowledge this is the first mixed-methods systematic review integrating quantitative and 482 qualitative evidence to explore the potential longer-term health outcomes of primary PPH. The 483 inclusion of qualitative evidence about women’s and partners’ experiences helped illuminate our 484 understanding of findings from the quantitative studies. A comprehensive and reproducible 485 search strategy was conducted, the methods were rigorous, and quality assessment and data 486 extraction were performed by two authors independently of each other. 487 Many of the limitations of this review reflect shortcomings of the included studies. First, there 488 were some challenges associated with drawing conclusions about the association between PPH 489 and some health outcomes because most included studies did not have appropriate comparison 490 groups and conclusions were made from the prevalence/proportions among the exposed group 491 without consideration of potential confounders. For most studies, the follow-up time was less 492 than five years. Most studies used self-reported health questionnaires, rather than objective 493 measurements, such as clinical diagnosis. 494 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 48 At the review level, the heterogeneity of studies, including different definitions of PPH, different 495 follow-up time and variation in the outcomes investigated, meant that it was not possible to 496 conduct meta-analysis. Additionally, the number of studies that made an attempt to establish the 497 potential longer-term impact of PPH was limited. Due to resource limitations, we were unable to 498 consider for inclusion five potentially eligible papers that were not written in English. 499 Implications for research 500 There is limited evidence about the longer-term impact of primary PPH on women and their 501 partners, hence there is a need for further research in this area. In particular, cohort studies with 502 comparison groups, and longer duration of follow-up should be conducted to explore the 503 association between PPH and selected chronic health outcomes, for instance, CVD. Study 504 designs could be improved by using standardised outcome assessment such as applying 505 standardised self-report tools validated for this population or clinical interviews. Additionally, 506 emphasis on the potential cumulative impact of PPH throughout the obstetric history of women 507 should be considered. This is particularly important in multiparous women who may have more 508 than one primary PPH with a potentially more severe impact. As evidenced by qualitative 509 studies, differences in quality of health care (e.g., frequency of follow up, access to counselling 510 and psychologist) should be also considered in any future prospective cohort studies. 511 In spite of increasing trends in PPH, women are less likely to experience adverse fatal outcomes 512 due to better management and recording of severe PPH compared to the past decade. (4, 82) 513 Globally, the research focus on other near-miss complications has shifted from maternal 514 mortality to surveillance of the burden of severe maternal morbidity, including improvement in 515 the quality of follow-up care. (83, 84) The same principle should be applied to future research on 516 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 49 PPH to explore whether the trajectory of PPH and its acute morbidities is associated with 517 subsequent risks of developing chronic disease. 518 More research should also be conducted to explore the potential longer-term psychological 519 implications for partners witnessing PPH. Qualitative evidence in this review suggested signs of 520 acute distress, with significant information needs and follow up support needs for partners who 521 witnessed women experiencing PPH. Obtaining information about partners’ mental health and 522 coping strategies could be beneficial to both the woman and her partner in terms of identifying 523 support needs moving forward. 524

525 This review synthesized evidence about longer-term physical and psychological health outcomes 526 among women who had a primary PPH in high-income countries, and their partners. While the 527 available evidence mainly focuses on immediate health outcomes following birth, this review 528 suggests that women can have longer-term health problems associated with PPH beyond one-529 year after childbirth. The extent of the impact of these health outcomes is poorly researched and 530 may be influenced by the severity of PPH, presence of other obstetric complications, and the 531 quality of care received. The limitations of the evidence about longer-term health outcomes after 532 PPH emphasizes the need for further research in this area. 533

534 We would like to express our sincere appreciation to Nia Roberts (Senior librarian, University of 535 Oxford Bodleian Health Care Libraries) for her guidance in developing search strategies. 536 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 50

537 1. WHO Recommendations for the Prevention and Treatment of Postpartum Haemorrhage. 538 WHO Guidelines Approved by the Guidelines Review Committee. Geneva2012. 539 2. Borovac-Pinheiro A, Pacagnella RC, Cecatti JG, Miller S, El Ayadi AM, Souza JP, et al. 540 Postpartum hemorrhage: new insights for definition and diagnosis. Am J Obstet Gynecol. 541 2018;219(2):162-8. 542 3. Knight M, Callaghan WM, Berg C, Alexander S, Bouvier-Colle M-H, Ford JB, et al. 543 Trends in Postpartum Hemorrhage in High Resource Countries: A Review and 544 Recommendations From the International Postpartum Hemorrhage Collaborative Group. 545 Obstetrical & Gynecological Survey. 2010;65(4):211-2. 546 4. Flood M, McDonald SJ, Pollock W, Cullinane F, Davey MA. Incidence, trends and 547 severity of primary postpartum haemorrhage in Australia: A population‐ based study using 548 Victorian Perinatal Data Collection data for 764 244 births. Australian and New Zealand Journal 549 of Obstetrics and Gynaecology. 2019;59(2):228-34. 550 5. Bateman BT, Berman MF, Riley LE, Leffert LR. The Epidemiology of Postpartum 551 Hemorrhage in a Large, Nationwide Sample of Deliveries. Obstetric Anesthesia Digest. 552 2011;31(2):89. 553 6. Meher S, Cuthbert A, Kirkham J, Williamson P, Abalos E, Aflaifel N, et al. Core 554 outcome sets for prevention and treatment of postpartum haemorrhage: an international Delphi 555 consensus study. BJOG: An International Journal of Obstetrics & Gynaecology. 2019;126(1):83-556 93. 557 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 51 7. Maswime S, Buchmann E. A systematic review of maternal near miss and mortality due 558 to postpartum hemorrhage. International Journal Of Gynaecology And Obstetrics. 559 2017;137(1):pp1-7. 560 8. Chauleur C, Cochery-Nouvellon E, Mercier E, Aya G, Marès P, Mismetti P, et al. 561 Analysis of the venous thromboembolic risk associated with severe postpartum haemorrhage in 562 the NOHA First cohort. Thrombosis and Haemostasis. 2008;100(5):773-9. 563 9. Hoveyda F, MacKenzie IZ. Secondary postpartum haemorrhage: incidence, morbidity 564 and current management. BJOG. 2001;108(9):927-30. 565 10. Liu S, Heaman M, Kramer MS, Demissie K, Wen SW, Marcoux S, et al. Length of 566 hospital stay, obstetric conditions at childbirth, and maternal readmission: a population-based 567 cohort study. Am J Obstet Gynecol. 2002;187(3):681-7. 568 11. Naz H, Sarwar I, Fawad A, Nisa AU. Maternal morbidity and mortality due to primary 569 PPH--experience at Ayub Teaching Hospital Abbottabad. J Ayub Med Coll Abbottabad. 570 2008;20(2):59-65. 571 12. Thompson JF, Heal LJ, Roberts CL, Ellwood DA. Women's breastfeeding experiences 572 following a significant primary postpartum haemorrhage: A multicentre cohort study. Int 573 Breastfeed J. 2010;5:5. 574 13. Thompson JF, Roberts CL, Ellwood DA. Emotional and physical health outcomes after 575 significant primary post-partum haemorrhage (PPH): a multicentre cohort study. The Australian 576 & New Zealand journal of obstetrics & gynaecology. 2011;51(4):365-71. 577 14. Carroll M, Daly D, Begley CM. The prevalence of women's emotional and physical 578 health problems following a postpartum haemorrhage: a systematic review. BMC Pregnancy 579 Childbirth. 2016;16:261. 580 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 52 15. Patricia E, Natasa K, Johanna L, Charlotte H, Linnea K, Ulf H, et al. Delineating the 581 Association between Heavy Postpartum Haemorrhage and Postpartum Depression. PLoS ONE. 582 2016;11(1):e0144274. 583 16. Ricbourg A, Gosme C, Gayat E, Ventre C, Barranger E, Mebazaa A. Emotional impact of 584 severe post-partum haemorrhage on women and their partners: an observational, case-matched, 585 prospective, single-centre pilot study. European Journal of Obstetrics and Gynecology. 586 2015;193:140-3. 587 17. van Steijn ME, Scheepstra KWF, Zaat TR, van Der Post JAM, Olff M, van Pampus MG. 588 Posttraumatic stress disorder in partners following severe postpartum haemorrhage: A 589 prospective cohort study. Women and birth : journal of the Australian College of Midwives. 590 2019. 591 18. Zaat TR, van Steijn ME, de Haan-Jebbink JM, Olff M, Stramrood CAI, van Pampus MG. 592 Posttraumatic stress disorder related to postpartum haemorrhage: A systematic review. Eur J 593 Obstet Gynecol Reprod Biol. 2018;225:214-20. 594 19. Pearson A, White H, Bath-Hextall F, Salmond S, Apostolo J, Kirkpatrick P. A mixed-595

approach to systematic reviews. Int J Evid Based Healthc. 2015;13(3):121-31. 596 20. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The 597 PRISMA 2020 statement: An updated guideline for reporting systematic reviews. Int J Surg. 598 2021;88:105906. 599 21. High Income countries classification: The World Bank; 2019 [Available from: 600 https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-601 lending-groups. 602 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 53 22. Bramer WM, Giustini D, de Jonge GB, Holland L, Bekhuis T. De-duplication of database 603 search results for systematic reviews in EndNote. J Med Libr Assoc. 2016;104(3):240-3. 604 23. Say L, Chou D, Gemmill A, Tunçalp Ö, Moller A-B, Daniels J, et al. Articles: Global 605 causes of maternal death: a WHO systematic analysis. The Lancet Global Health. 2014;2:e323-606 e33. 607 24. Kim SY, Park JE, Lee YJ, Seo HJ, Sheen SS, Hahn S, et al. Testing a tool for assessing 608 the risk of bias for nonrandomized studies showed moderate reliability and promising validity. J 609 Clin Epidemiol. 2013;66(4):408-14. 610 25. Critical Appraisal Skills Programme (2018) 2018 [Available from: https://casp-611 uk.net/wp-content/uploads/2018/01/CASP-Qualitative-Checklist-2018.pdf. 612 26. Greenhalgh T, Robert G, Macfarlane F, Bate P, Kyriakidou O, Peacock R. Storylines of 613 research in diffusion of innovation: a meta-narrative approach to systematic review. Soc Sci 614 Med. 2005;61(2):417-30. 615 27. Coates R, Cupples G, Scamell A, McCourt C. Women's experiences of induction of 616 labour: Qualitative systematic review and thematic synthesis. Midwifery. 2019;69:17-28. 617 28. Walker RC, Graham A, Palmer SC, Jagroop A, Tipene-Leach DC. Understanding the 618 experiences, perspectives and values of indigenous women around smoking cessation in 619 pregnancy: systematic review and thematic synthesis of qualitative studies. Int J Equity Health. 620 2019;18(1):74. 621 29. Thomas J, Harden A. Methods for the thematic synthesis of qualitative research in 622 systematic reviews. BMC Med Res Methodol. 2008;8:45. 623 30. Lucylynn Lizarondo1 CS, Judith Carrier, Christina Godfrey, Kendra Rieger, Susan 624 Salmond, Joao Apostolo, Pamela Kirkpatrick, Heather Loveday. Chapter 8: Mixed methods 625 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 54 systematic reviews. 2017. In: Joanna Briggs Institute Reviewer's Manual [Internet]. Joanna 626 Briggs Institute. Available from: https://reviewersmanual.joannabriggs.org/. 627 31. Chauleur C, Cochery-Nouvellon E, Mercier E, Aya G, Mares P, Mismetti P, et al. 628 Analysis of the venous thromboembolic risk associated with severe postpartum haemorrhage in 629 the NOHA First cohort. Thrombosis & Haemostasis. 2008;100(5):773-9. 630 32. Chessman J, Patterson J, Nippita T, Drayton B, Ford J. Haemoglobin concentration 631 following postpartum haemorrhage and the association between blood transfusion and 632 breastfeeding: a retrospective cohort study. BMC Research Notes. 2018;11(1):686. 633 33. Drayton BA, Patterson JA, Nippita TA, Ford JB. Red blood cell transfusion after 634 postpartum haemorrhage and breastmilk feeding at discharge: A population-based study. 635 Australian & New Zealand Journal of Obstetrics & Gynaecology. 2016;56(6):591-8. 636 34. Eckerdal P, Kollia N, Lofblad J, Hellgren C, Karlsson L, Hogberg U, et al. Delineating 637 the Association between Heavy Postpartum Haemorrhage and Postpartum Depression. PLoS 638 ONE [Electronic Resource]. 2016;11(1):e0144274. 639 35. Thurn L, Wikman A, Lindqvist PG. Postpartum blood transfusion and hemorrhage as 640 independent risk factors for venous thromboembolism. Thrombosis Research. 2018;165:54-60. 641 36. van Steijn ME, Scheepstra KWF, Zaat TR, van der Post JAM, Olff M, van Pampus MG. 642 Posttraumatic stress disorder in partners following severe postpartum haemorrhage: A 643 prospective cohort study. Women & Birth: Journal of the Australian College of Midwives. 644 2019;13:13. 645 37. van Steijn ME, Scheepstra KWF, Zaat TR, van Rooijen DE, Stramrood CAI, Dijksman 646 LM, et al. Severe postpartum hemorrhage increases risk of posttraumatic stress disorder: a 647 prospective cohort study. J Psychosom Obstet Gynaecol. 2020:1-11. 648 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 55 38. Briley AL, Silverio SA, Singh C, Sandall J, Bewley S. "It's like a bus, going downhill, 649 without a driver": A qualitative study of how postpartum haemorrhage is experienced by women, 650 their birth partners, and healthcare professionals. Women Birth. 2020. 651 39. Knight M, Acosta C, Brocklehurst P, Cheshire A, Fitzpatrick K, Hinton L, et al. NIHR 652 Journals Library Programme Grants for Applied Research. 2016;06:06. 653 40. Thompson JF, Roberts CL, Ellwood DA. Emotional and physical health outcomes after 654 significant primary post-partum haemorrhage (PPH): a multicentre cohort study. Australian & 655 New Zealand Journal of Obstetrics & Gynaecology. 2011;51(4):365-71. 656 41. Thompson JF, Ford JB, Raynes-Greenow CH, Roberts CL, Ellwood DA. Women's 657 experiences of care and their concerns and needs following a significant primary postpartum 658 hemorrhage. Birth. 2011;38(4):327-35. 659 42. Liu C, Butwick A, Sand A, Wikstrom AK, Snowden JM, Stephansson O. The association 660 between postpartum hemorrhage and postpartum depression: A Swedish national register-based 661 study. PLoS One. 2021;16(8):e0255938. 662 43. de la Cruz CZ, Coulter ML, O'Rourke K, Amina Alio P, Daley EM, Mahan CS. Women's 663 experiences, emotional responses, and perceptions of care after emergency peripartum 664 hysterectomy: a qualitative survey of women from 6 months to 3 years postpartum. Birth. 665 2013;40(4):256-63. 666 44. de la Cruz CZ, Coulter M, O'Rourke K, Mbah AK, Salihu HM. Post-traumatic stress 667 disorder following emergency peripartum hysterectomy. Archives of Gynecology & Obstetrics. 668 2016;294(4):681-8. 669 45. Feinberg EC, Molitch ME, Endres LK, Peaceman AM. The incidence of Sheehan's 670 syndrome after obstetric hemorrhage. Fertility & Sterility. 2005;84(4):975-9. 671 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 56 46. Dunning T, Harris JM, Sandall J. Women and their birth partners' experiences following 672 a primary postpartum haemorrhage: a qualitative study. BMC Pregnancy & Childbirth. 673 2016;16:80. 674 47. Michelet D, Ricbourg A, Gosme C, Rossignol M, Schurando P, Barranger E, et al. 675 Emergency hysterectomy for life-threatening postpartum haemorrhage: Risk factors and 676 psychological impact. Gynecologie, Obstetrique & Fertilite. 2015;43(12):773-9. 677 48. Sentilhes L, Gromez A, Clavier E, Resch B, Descamps P, Marpeau L. Long-term 678 psychological impact of severe postpartum hemorrhage. Acta Obstet Gynecol Scand. 679 2011;90(6):615-20. 680 49. van Stralen G, Ruijten LLM, Kaptein AA, Wolterbeek R, van Roosmalen J. Major 681 obstetric hemorrhage: a follow-up survey on quality of life of women and their partners. Journal 682 of Psychosomatic Obstetrics & Gynecology. 2018;39(1):38-46. 683 50. Ukah UV, Platt RW, Potter BJ, Paradis G, Dayan N, He S, et al. Obstetric haemorrhage 684 and risk of cardiovascular disease after three decades: a population-based cohort study. BJOG. 685 2020. 686 51. Cho GJ, Lee KM, Kim HY, Han SW, Oh MJ, Chiec L, et al. Postpartum haemorrhage 687 requiring transfusion and risk of cardiovascular disease later in life: a retrospective cohort study. 688 BJOG. 2021;128(4):738-44. 689 52. Parry-Smith W, Sumilo D, Subramanian A, Gokhale K, Okoth K, Gallos I, et al. 690 Postpartum haemorrhage and risk of long-term hypertension and cardiovascular disease: an 691 English population-based longitudinal study using linked primary and secondary care databases. 692 BMJ Open. 2021;11(5):e041566. 693 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 57 53. Parry-Smith W, Okoth K, Subramanian A, Gokhale KM, Chandan JS, Humpston C, et al. 694 Postpartum haemorrhage and risk of mental ill health: A population-based longitudinal study 695 using linked primary and secondary care databases. J Psychiatr Res. 2021;137:419-25. 696 54. Eggel B, Bernasconi M, Quibel T, Horsch A, Vial Y, Denys A, et al. Gynecological, 697 reproductive and sexual outcomes after uterine artery embolization for post-partum haemorrage. 698 Sci Rep. 2021;11(1):833. 699 55. Elmir R, Schmied V, Wilkes L, Jackson D. Separation, failure and temporary 700 relinquishment: women's experiences of early mothering in the context of emergency 701 hysterectomy. Journal of Clinical Nursing. 2012;21(7-8):1119-27. 702 56. Elmir R, Schmied V, Jackson D, Wilkes L. Between life and death: women's experiences 703 of coming close to death, and surviving a severe postpartum haemorrhage and emergency 704 hysterectomy. Midwifery. 2012;28(2):228-35. 705 57. Elmir R, Jackson D, Schmied V, Wilkes L. 'Less feminine and less a woman': the impact 706 of unplanned postpartum hysteretomy on women. International Journal of Childbirth. 707 2012;2(1):51-60. 708 58. Elmir R. Finding Meaning in Life Following Emergency Postpartum Hysterectomy: 709 What Doesn't Kill Us Makes Us Stronger. J Midwifery Womens Health. 2014;59(5):510-5. 710 59. Woiski MD, Belfroid E, Liefers J, Grol RP, Scheepers HC, Hermens RP. Influencing 711 factors for high quality care on postpartum haemorrhage in the Netherlands: patient and 712 professional perspectives. BMC Pregnancy & Childbirth. 2015;15:272. 713 60. Bernasconi M, Eggel-Hort B, Horsch A, Vial Y, Denys A, Quibel T, et al. Paternal and 714 maternal long-term psychological outcomes after uterine artery embolization for severe post-715 partum hemorrhage. Scientific reports. 2021;11(1):13990-. 716 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 58 61. Snowdon C, Elbourne D, Forsey M, Alfirevic Z. Information-hungry and disempowered: 717 a qualitative study of women and their partners' experiences of severe postpartum haemorrhage. 718 Midwifery. 2012;28(6):791-9. 719 62. Elmir R. Finding Meaning in Life Following Emergency Postpartum Hysterectomy: 720 What Doesn't Kill Us Makes Us Stronger. Journal of Midwifery & Women's Health. 721 2014;59(5):510-5. 722 63. de Visser SM, Kirchner CA, van der Velden BGJ, de Wit AC, Dijkman A, Huisjes AJM, 723 et al. Major obstetric hemorrhage: Patients' perspective on the quality of care. European Journal 724 of Obstetrics, Gynecology, & Reproductive Biology. 2018;224:146-52. 725 64. Ayers S, Eagle A, Waring H. The effects of childbirth-related post-traumatic stress 726 disorder on women and their relationships: A qualitative study. Psychology, health & medicine. 727 2006;11(4):389-98. 728 65. Peeler S, Stedmon J, Chung MC, Skirton H. Women's experiences of living with 729 postnatal PTSD. Midwifery. 2018;56:70-8. 730 66. Beck CT. Post-Traumatic Stress Disorder Due to Childbirth: The Aftermath. Nurs Res. 731 2004;53(4):216-24. 732 67. Yildiz PD, Ayers S, Phillips L. The prevalence of posttraumatic stress disorder in 733 pregnancy and after birth: A systematic review and meta-analysis. J Affect Disord. 734 2017;208:634-45. 735 68. Dekel S, Stuebe C, Dishy G. Childbirth Induced Posttraumatic Stress Syndrome: A 736 Systematic Review of Prevalence and Risk Factors. Front Psychol. 2017;8:560. 737 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 59 69. Karpati PCJ, Rossignol M, Payen D, Mebazaa A, Pirot M, Cholley B, et al. High 738 incidence of myocardial ischemia during postpartum hemorrhage. Anesthesiology. 739 2004;100(1):30-6. 740 70. Arnott C, Patel S, Hyett J, Jennings G, Woodward M, Celermajer DS. Women and 741 Cardiovascular Disease: Pregnancy, the Forgotten Risk Factor. Heart Lung Circ. 742 2020;29(5):662-7. 743 71. Bhattacharya S, Prescott GJ, Iversen L, Campbell DM, Smith WC, Hannaford PC. 744 Hypertensive disorders of pregnancy and future health and mortality: A record linkage study. 745 Pregnancy Hypertens. 2012;2(1):1-7. 746 72. Auger N, Potter BJ, He S, Healy-Profitos J, Schnitzer ME, Paradis G. Maternal 747 Cardiovascular Disease 3 Decades After Preterm Birth: Longitudinal Cohort Study of Pregnancy 748 Vascular Disorders. Hypertension. 2020;75(3):788-95. 749 73. Azulay CE, Pariente G, Shoham-Vardi I, Kessous R, Sergienko R, Sheiner E. Maternal 750 anemia during pregnancy and subsequent risk for cardiovascular disease. J Matern Fetal 751 Neonatal Med. 2015;28(15):1762-5. 752 74. Li J, Song C, Li C, Liu P, Sun Z, Yang X. Increased risk of cardiovascular disease in 753 women with prior gestational diabetes: A systematic review and meta-analysis. Diabetes Res 754 Clin Pract. 2018;140:324-38. 755 75. Reed R, Sharman R, Inglis C. Women's descriptions of childbirth trauma relating to care 756 provider actions and interactions. BMC Pregnancy Childbirth. 2017;17(1):21. 757 76. Elmir R, Schmied V, Wilkes L, Jackson D. Women's perceptions and experiences of a 758 traumatic birth: a meta-ethnography. J Adv Nurs. 2010;66(10):2142-53. 759 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 60 77. Kirova K, Snell T. Women's experiences of positive postnatal partner support. J Reprod 760 Infant Psychol. 2019;37(2):206-18. 761 78. Etheridge J, Slade P. "Nothing's actually happened to me.": the experiences of fathers 762 who found childbirth traumatic. BMC pregnancy and childbirth. 2017;17(1):80. 763 79. Daniels E, Arden-Close E, Mayers A. Be quiet and man up: a qualitative questionnaire 764 study into fathers who witnessed their Partner's birth trauma. BMC pregnancy and childbirth,. 765 2020;20(1):236. 766 80. Darwin Z, Galdas P, Hinchliff S, Littlewood E, McMillan D, McGowan L, et al. Fathers' 767 views and experiences of their own mental health during pregnancy and the first postnatal year: a 768 qualitative interview study of men participating in the UK Born and Bred in Yorkshire (BaBY) 769 cohort. BMC Pregnancy Childbirth. 2017;17(1):45. 770 81. Vallin E, Nestander H, Wells MB. A literature review and meta-ethnography of fathers' 771 psychological health and received social support during unpredictable complicated childbirths. 772 Midwifery. 2019;68:48-55. 773 82. Ford JB, Patterson JA, Seeho SK, Roberts CL. Trends and outcomes of postpartum 774 haemorrhage, 2003-2011. BMC Pregnancy Childbirth. 2015;15:334. 775 83. Knight M, Callaghan WM, Berg C, Alexander S, Bouvier-Colle MH, Ford JB, et al. 776 Trends in postpartum hemorrhage in high resource countries: a review and recommendations 777 from the International Postpartum Hemorrhage Collaborative Group. BMC Pregnancy 778 Childbirth. 2009;9:55. 779 84. Geller SE, Koch AR, Garland CE, MacDonald EJ, Storey F, Lawton B. A global view of 780 severe maternal morbidity: moving beyond maternal mortality. Reprod Health. 2018;15(Suppl 781 1):98. 782 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint 61 Supporting Information 783 S1 Checklist. PRISMA 2020 checklist. 784 S1 File. Medline Search Strategy. 785 S2 File. Data extraction forms 786 S1 Table. Modified risk of bias assessment tool for non-randomized studies (ROBANS) 787 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 25, 2022. ; https://doi.org/10.1101/2022.08.24.22279096doi: medRxiv preprint