{"paper_id":"4066f1b4-a62e-494f-876a-f3727f49fbe5","body_text":"1 \nPreterm Prelabour Rupture Of Membranes (PPROM) before 23 weeks \ngestation: A prospective observational study \n \nGoodfellow, L1, Care, A1, Curran, C2, Roberts, D3, Turner, M.A.1,3, Knight, M4, Alfirevic, Z.1  \n1Harris-Wellbeing Research Centre, University of Liverpool.  \n2Little Heartbeats Patient Support Group \n3Liverpool Women’s NHS Foundation Trust \n4National Perinatal Epidemiology Unit, University of Oxford \n \nABSTRACT \n \nObjectives  \nDescribe infant and maternal outcomes of a national cohort of women with preterm \nprelabour rupture of membranes (PPROM) under 23 weeks gestation. \n \nDesign  \nProspective national population-based cohort study using the UK Obstetric Surveillance \nSystem (UKOSS). \n \nSetting  \nAll 194 obstetric units in the UK. \n \nParticipants  \n330 women with singleton and 38 with multiple pregnancies and PPROM between 16+0 and \n22+6 weeks gestation 1/9/19-28/2/21. \n \nMain outcome measures \nInfant outcomes: livebirth, survival to hospital discharge and severe morbidity, defined as \nintraventricular haemorrhage grade 3 or 4 and/or supplemental oxygen requirement at 36 \nweeks postmenstrual age.  \nMaternal outcomes: surgery for placental removal; sepsis; admission to intensive treatment \nunit (ITU) and death.  \n \nMethods \nAll data including rates of termination of pregnancy for medical reasons (TFMR) were \nreported. Three rates were calculated for infant outcomes: i) all TFMR excluded; ii) \nassuming that all TFMR and those with missing data would have died; iii) assuming that all \nTFMR and those with missing data would be liveborn. Rates are presented as i (ii to iii). \n \nResults \nFor singleton pregnancies the livebirth rate was 44% (30 to 62%), infant survival to discharge \nwas 26% (16 to 54%) and 18% (12 to 49%) of infants survived without severe morbidity. \nMaternal sepsis rate was 12% for singleton and 26% for twin pregnancies. Surgery for \nplacental removal was 20% and 14%, respectively.  \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.\n\n   2 \nFive women became severely unwell with sepsis, 2 died and a further 3 required ITU care. \n \nConclusions \nAlthough significant numbers of pregnancies with very early PPROM have favourable \noutcomes, morbidity and mortality rates in this cohort are high for mothers and infants. \n \nThese data can be used in counselling families facing PPROM prior to 23 weeks gestation \nand to underpin research into the complex pathologies, including sepsis, related to this \ncondition. Currently available guidelines should be updated accordingly. \n \nWhat is already known on this topic \n \n• PPROM under 23 weeks gestation is a serious pregnancy complication with high \nrates of morbidity for mothers and infants \n• Women are often advised to consider termination for medical reasons (TFMR) \n• Contemporary, population based, pregnancy outcomes are not available, making \ncounselling even more difficult  \n \nWhat this study adds \n \n• This study identified significant maternal morbidity; 12% of women developed sepsis \nand 2 women (0.6%, 95%CI 0.17-2.2%) died \n• Conversely infant outcomes were relatively favourable; 26% of expectantly managed \ninfants survived to hospital discharge and the potential worst-best case survival \nrange including those that had termination for medical reasons (TFMR) was 16-54% \n• Understanding of these results are imperative to appropriate counselling and \nmanagement of women facing this difficult complication \n \nIntroduction \n \nPreterm prelabour rupture of membranes (PPROM) complicates 30-40% of all preterm \nbirths. [1] Serious complications of PPROM include chorioamnionitis leading to maternal \nand/or neonatal sepsis, placental abruption and stillbirth. UK clinical guidelines exist for \nmanagement of this condition but only for pregnancies after 24 weeks gestation, i.e. once \npregnancy is legally viable.[1] Prior to this gestation, the burden of neonatal morbidity and \nmortality has been considered so high that termination of pregnancy is generally offered \ndue to extremely low fetal survival and concerns about lifelong neurological disability \nsecondary to extreme prematurity.[2] The incidence of PPROM below 23 weeks gestation is \nlow (~0.1%), therefore, a typical obstetric unit with 4000 births a year will manage fewer \nthan 5 cases annually.[3] This has led to both a paucity of research and a paucity of clinical \nexperience in expectant management of this condition.   \n \nWomen and their families report that clinical care in the UK, including offer of termination \nof pregnancy for medical reasons (TFMR) differs broadly in seemingly similar clinical \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   3 \nscenarios.[4] These inconsistencies understandably add to parental distress under already \ndifficult circumstances.  \n \nThe aim of this study was to provide UK population level data for pregnancies with PPROM \nbetween 16+0 and 22+6 weeks gestation, stratified according to gestation when PPROM \noccurred.  The study was carried out using the UK Obstetric Surveillance System (UKOSS), a \nresearch infrastructure that encompasses every consultant-led maternity department in the \ncountry.  \nMethod \nData collection \nThe UK Obstetric Surveillance System (UKOSS) is a research platform that collects \npopulation-based information about rare pregnancy events from all 194 consultant-led \nmaternity hospitals in the UK. [5]  \n \nNominated reporting clinicians notified UKOSS of all pregnant women who experienced \npreterm prelabour rupture of membranes between 16+0 and 22+6 weeks gestation \n(inclusive). The two exclusion criteria were: pregnancies in which the membranes ruptured \nbefore 16+0 weeks gestation but were only diagnosed in the 16+0 and 22+6 week period; and \ncases where intrauterine death of all fetuses was diagnosed before rupture of membranes. \nTo capture all relevant pregnancies, the minimum latency between PPROM and labour or \nbirth was not specified.  Information was requested about all reported pregnancies using a \nset proforma [6] and regular reminders to return missing data were sent at weeks 6, 10, 14 \nand 28 after notification, and a final reminder at the end of the data collection period in \nSeptember 2021. If a woman was still pregnant when the initial data collection form was \nreceived then pregnancy outcome data were requested at 2, 6, 10, 14 and 28 weeks after \nthe estimated due date.  Referring hospitals that the woman or infant were transferred to \nwere also contacted to request outcome data. \n \nWe planned to collect data on pregnancies with PPROM from 1st September 2019 to 31st \nAugust 2020, however, when the COVID-19 pandemic was declared in the UK in March 2020 \nthe study was extended for six-months to investigate potential changes in outcomes \nsecondary to the pandemic. Therefore, the study included women with PPROM between 1st \nSeptember 2019 and 28th February 2021 (inclusive).  \n \nSample size and statistical analysis \nAs this was a time limited national observational study no formal power calculation was \ncarried out. The incidence rate was calculated using the denominator of maternities in 2020 \nfrom the constituent nations of the UK.[7–9] Statistical analysis was performed in Stata \nversion 15.1 (StataCorp). The study is reported in accordance with the Strengthening the \nreporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for \nreporting observational studies.[10]  \n \nDemographics are reported for the whole cohort and divided according to whether the \npregnancy had expectant management or TFMR. Women who had TFMR after a period of \nexpectant management were included within the TFMR group. Gestational age at PPROM \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   4 \nand at birth were calculated according to ultrasound assessment of estimated date of \ndelivery. \n \nMaternal age at estimated date of delivery was calculated assuming a date of birth of 1st \nJune within the given year of birth since only maternal year of birth was collected to \nmaintain anonymity. Body mass index was based on first recorded weight and height in \npregnancy. Ethnicity was recorded from medical records, based on the woman’s self-report. \nAdverse pregnancy history was noted if a woman had previous pregnancy with PPROM \nbetween 16+0 and 33+6 weeks gestation, midtrimester loss between 16+0 and 22+6 weeks \ngestation or spontaneous preterm birth (PTB) between 23+0 and 36+6 weeks. \n \nThe impact of the COVID-19 pandemic on pregnancy outcomes was assessed by grouping \nsingleton pregnancies into those with PPROM between 1st September 2019 and 29th \nFebruary 2020 (‘prior to COVID-19’) and between 1st March 2020 and 28th February 2021 \n(‘during COVID-19’ pandemic). The rate of reported pregnancies and infant and maternal \npregnancy outcomes were compared across the two groups. \n \nThe calculated latency between PPROM and birth included all women except those that had \na TFMR, thereby including women who had spontaneous births, intrauterine deaths and \nmedically indicated births. \n \nPregnancy outcomes for infants and mothers are reported separately for singleton and twin \npregnancies. Higher order multiples are briefly described. Singleton pregnancy outcomes \nwere divided into four mutually exclusive groups based on when PPROM occurred: 16+0-17+6 \nweeks gestation; 18+0-19+6 weeks gestation; 20+0-21+6 weeks gestation; and 22+0-22+6 weeks \ngestation. The group comprising 22+0-22+6 weeks gestation was analysed separately because \nthe British Association of Perinatal Medicine (BAPM) produced guidelines in October 2019 \nsuggesting that in some pregnancies, with parental agreement, active resuscitation should \nbe considered at birth from 22+0 weeks gestation.[11] Twin pregnancy outcomes are \npresented according to chorionicity of the pregnancy. \n \nInfant outcomes were: livebirth and survival to hospital discharge with or without severe \nmorbidity defined as intraventricular haemorrhage grade 3 or 4 and/or requirement for \nsupplemental oxygen at 36 weeks postmenstrual age. This outcome was selected in order to \nallow compatibility with the study by Kibel et al. [13] In addition reporting clinicians were \nasked to record whether the infant had limb contractures, neonatal seizures, congenital \nanomalies or severe lung disease during the neonatal course. Severe lung disease was \ndefined as requiring high frequency oscillatory ventilation during the neonatal admission, \ninhaled nitric oxide during the neonatal admission, or supplemental oxygen at 36 weeks \npostmenstrual age. \n \nTo account for the impact of termination of pregnancy for medical reasons (TFMR) and \nmissing data on the calculated rates of infant outcomes, three rates were calculated: (1) \ninfant outcome in expectantly managed pregnancies with known infant outcome; (2) infant \noutcome assuming that all pregnancies that had TFMR or an unknown outcome had died; \nand (3) infant outcome assuming all pregnancies that had TFMR or an unknown outcome \nhad survived.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   5 \n \nDetails of the type of pregnancy loss that occurred are presented by allocating infants into \none of five mutually exclusive groups: birth or intrauterine death before 22+0 weeks \ngestation (often called miscarriage); intrauterine death 22+0 or more weeks gestation; \nneonatal death; TFMR without expectant management; and TFMR after a period of \nexpectant management.  \n \nMaternal outcomes were sepsis, surgery for placental removal, intensive treatment unit \n(ITU) admission and death. All maternal outcomes are reported as defined by local \nclinicians. \n \nData are presented as descriptive statistics (mean/median and standard deviation/IQR) and \ndifferences between groups were compared using t test for maternal age, Mann-Whitney U \ntest for gestational age at PPROM and Chi squared tests for categorical variables, except for \nmaternal death and ITU admission according to COVID-19 pandemic status which were \ncompared using Fisher’s exact tests. The Wilson score interval was used to generate \nconfidence intervals for proportions where appropriate. A p value <0.05 was considered \nstatistically significant. \n \nEthics statement  \nEthics committee approval was obtained from the North London REC1 (Ref. Number \n10/H0717/20). Cases were reported anonymously by nominated hospital reporting \nclinicians, and as such consent from patients was neither required nor sought. Further \ninformation is available at https://www.npeu.ox.ac.uk/ukoss/completed-\nsurveillance/epprom. \n \nPatient and public involvement \nThe patient support and advocacy group, Little Heartbeats, approached the author AC with \nconcerns about inconsistency in counselling and management of cases of PPROM prior to 23 \nweeks gestation, stimulating this research. CC (the founder of Little Heartbeats) and the \npatient and public members of the UKOSS Steering Committee were then involved in the \ndesign of the study, the conduct of the study and interpretation of the results. CC met \nregularly with authors LG and AC to review the findings and plan the optimal presentation \nof the data. The completed analysis was also reviewed by patient and public representation \nwithin the UKOSS Steering Committee. \n \nResults \n \nAll 194 UK hospitals participated in UKOSS. One hundred and twenty five (64%) hospitals \nreported at least one woman with PPROM at 16+0-22+6 weeks gestation, leading to 551 \nwomen reported in total. One hundred and seventy-nine women were removed due to \nduplication, ineligibility or insufficient information to assess eligibility. Four women were \nexcluded due to false positive diagnosis of PPROM, defined as women in which all follow-up \nultrasound scans showed normal liquor volume and birth was at or after 37+0 weeks \ngestation. \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   6 \nIn total 368 women were included in the analysis of whom 38 had multiple pregnancies.  \nThere were an estimated 1 011 924 maternities in the UK over 18 months.[7–9] The \nestimated incidence is therefore 1 case per 2750 maternities (0.04%). \n \nDemographics \nMaternal characteristics were similar when comparing women who had expectant \nmanagement and those with TFMR (Table 1). Women that had TFMR had an earlier median \ngestation of PPROM (Table 1). \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   7 \n \n \nTable 1: Demographics of cohort, presented both for the whole cohort and according to whether the woman had expectant management or Termination For Medical Reasons (TFMR).  \np value\n32 (6) 32 (6) 33 (5) 0.198\n82 (22%) 62 (24%) 20 (18%) 0.150\n51 (14%) 36 (14%) 15 (13%) 0.860\nAsian 62 (17%) 42 (16%) 20 (18%)\nBlack 36 (10%) 24 (9%) 12 (11%)\nMixed/any other ethnic group 14 (4%) 10 (4%) 4 (4%)\nWhite 250 (68%) 173 (68%) 77 (68%)\nNot specified 6 (2%) 6 (2%) 0 (0%)\nPrimiparous 157 (43%) 106 (42%) 51 (45%) 0.584\nAt least one previous pregnancy affected by any \nof: PPROM/midtrimester loss/preterm birth 63 (17%) 40 (16%) 23 (20%) 0.273\n330 (90%) 227 (89%) 103 (91%) 0.535\n19+3 (17+6-21+1) 19+6 (18+1-21+2) 18+6 (17+2-20+2) 0.000\nMaternal BMI over 35 (n, %)\nMaternal \nethnicity    (n, \n%)\n0.579\nObstetric \nhistory (n, %)\nSingleton pregnancy (n, %)\nGestational age at PPROM (median, IQR)\nTotal n 368 255 113\nMaternal age in years (mean, SD)\nMaternal smoking at booking appointment (n, %)\nWhole cohort\nManagement group Comparison of \nmanagement \ngroups Expectant Termination For Medical \nReasons (TFMR)\n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   8 \nCOVID-19 Pandemic \nThere were 140 women with PPROM in singleton pregnancies reported to the study in the 6 \nmonths prior to the COVID-19 pandemic and 190 in the year during the pandemic. The \nincidence of reported cases of PPROM was higher prior to the pandemic compared to \nduring the pandemic, with a median of 23 (IQR 19.5-24.5) cases/month and 16.5 (IQR 14-\n20.25) cases/month respectively (p=0.0004). The number of reported cases per month was \nlowest for the period July 2020-December 2020 (Figure A1). There were no significant \ndifferences detected in infant or maternal pregnancy outcomes according to whether the \nPPROM occurred prior to or during the pandemic (Tables A1 and A2), therefore the \nremainder of the analysis was performed using the whole dataset. \n \nPregnancy after PPROM \nIt was possible to calculate the latency between PPROM and birth in 223/227 (98%) of \nexpectantly managed singleton pregnancies. In the immediate period after PPROM the \nchance of birth was high: 27% (60/223) of births occurred within 72 hours of PPROM and a \nfurther 12% (27/223) by 7 days after PPROM (Table 2). Amongst those women that \nremained pregnant the chance of birth was 21% (29/136) in the second week after PPROM \nand from the third week onwards the chance of giving birth was approximately 16% of \nthose who remained pregnant per week. This pattern was consistent across the gestational \nages of PPROM studied (Figure A2). \n \n \nTable 2: Latency between PPROM and birth in singleton pregnancies with expectant management. Pregnancies that had a \ntermination of pregnancy (n=103) and pregnancies with unknown livebirth status (n=4), are excluded from this table. p \nvalue compares latency by gestational age category at PPROM and are calculated by chi squared test. \n \nInfant outcomes \n \nThe rate of TFMR declined as gestation of PPROM advanced, from 46% (39/84) of women \nwith singleton pregnancies with PPROM at 16+0 -17+6 weeks to 19% (7/37) at 22+0-22+6 \nweeks (p=0.004, Table 3).  \n \nAmongst the singleton pregnancies that had expectant management and known infant \noutcome the overall rate of livebirth was 44% (98/223) (Table 3). If one assumes that all \npregnancy with TFMR and those with missing data could have been liveborn, the livebirth \nrange would be 62%. In the worst-case scenario with no livebirths in the cohort of \npregnancies with TFMR or missing data the livebirth rate estimate is only 30%.  \n \nGestational age at PPROM is an important confounder in expectantly managed pregnancies \nwith the livebirth rate rising from 33% (14/43) in pregnancies with PPROM at 16+0 -17+6 \nweeks to 67% (20/30) in pregnancies with PPROM at 22+0-22+6 weeks (p=0.023) (Table 3).  \n \nLess than 72 hours 60 27% 16 (37%) 18 (26%) 20 (25%) 6 (20%)\n72 hours to <7 days 27 12% 4 (9%) 8 (11%) 9 (11%) 6 (20%)\n7 days to <28 days 48 22% 6 (14%) 12 (17%) 24 (30%) 6 (20%)\n28 days or more 85 38% 17 (40%) 32 (46%) 26 (33%) 10 (33%)\nNot specified 3 1% 0 (0%) 0 (0%) 1 (1%) 2 (7%)\n0.303\nn \nSingleton pregnancies with \nexpectant management\n70 80 30223\np value\nLatency \nbetween \nPPROM and \nbirth\nGestation at PPROM (weeks)\n22+0-22+620+0-21+618+0-19+616+0-17+6\n43\nWhole cohort\n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   9 \nThe overall rate of infant survival to discharge with expectant management was 26% \n(54/207) when all data needed to assess this outcome were available. The range-based \nassumption related to TFMR and missing data was 54% (177/330) for the best-case scenario \nand 16% (54/330) for the worst-case scenario (all TFMR and pregnancies with missing data \nhaving an adverse outcome). There was a trend towards improved survival with advancing \ngestational age at PPROM, but this did not reach our threshold for statistical significance \n(p=0.265) (Table 3). \n \nThe rate of survival to discharge amongst liveborn infants was 55% (54/98). A further 20% \n(20/98) of liveborn infants had missing data about their discharge status. \n \nThe median gestation at birth of liveborn infants was 28+3 weeks, interquartile range (IQR) \n25+3-30+2 and the median gestation at birth of infants with known survival to discharge was \n29+4 weeks, IQR 27+1-34+3 weeks. The median length of hospital stay after birth for surviving \ninfants was 59 days, IQR 17-100 days. \n \n \nTable 3: Infant outcomes for singleton pregnancies. p values compare outcomes by gestational age category at PPROM and \nare calculated by chi squared test. \n^Missing data by gestation at PPROM: 16+0-17+6 weeks livebirth status (2) and discharge status (5); 18+0-19+6 weeks \nlivebirth status (0) and discharge status (3); 20+0-21+6 weeks livebirth status (2) and discharge status (8); 22+0-22+6 weeks \nlivebirth status (0) and discharge status (4) \n*Excludes pregnancies with termination for medical reasons (TFMR) and those with livebirth status missing \n** Excludes pregnancies with TFMR and those with discharge status missing \nAmongst the 54 singleton liveborn neonates with known survival to hospital discharge, 16 \n(30%) met our criteria for severe morbidity with grade 3 or 4 IVH and/or requirement for \nsupplemental oxygen therapy at 36 weeks postmenstrual age (Table 3). The rate of survival \nwithout severe morbidity amongst all expectantly managed singleton pregnancies with \nknown outcomes was 18% (38/207).  The best-case scenario if infants of all pregnancies \nwith TFMR and missing data had favourable outcomes is a survival to discharge without \nsevere morbidity of 49% (161/330) and the worst-case scenario if all infants of pregnancies \nwith TFMR and missing data died is a survival to discharge without severe morbidity rate of \n12% (38/330) (Table 3).  \n \nThe infant morbidity included grade 3 or 4 IVH in 11% (6/54) of survivors and severe lung \ndisease after birth was reported in 52% (28/54) of surviving infants. Two neonates had limb \ncontractures, affecting one and two limbs respectively and no surviving infants had neonatal \nseizures. There was no difference in morbidity according to gestational age at PPROM, \nhowever the small number of surviving infants with earlier gestations of PPROM limited \nstatistical power.  The rate of survival without severe morbidity appeared to be influenced \nSingleton pregnancies\nLivebirth after expectant management* 98/223 44% 14/43 33% 27/70 39% 37/80 46% 20/30 67% 0.023\nRange based on all singletons (30 -62%) (17 -65%) (26 -58%) (35 -58%) (54 -73%)\nSurvival to hospital discharge after expectant management** 54/207 26% 7/40 18% 16/67 24% 21/74 28% 10/26 38% 0.265\nRange based on all singletons (16 -54%) (8 -61%) (16 -66%) (20 -49%) (27 -57%)\nSuvival without severe morbidity after expectant management** 38/207 18% 5/40 13% 11/67 16% 13/74 59% 9/26 35% 0.127\nRange based on all singletons (12 -49%) (6 -52%) (11 -45%) (12 -41%) (24 -54%)\nTermination For Medical Reasons (TFMR) 103/330 31% 39/84 46% 32/102 31% 25/107 23% 7/37 19% 0.004\nWhole cohort\n330^\np value16+0-17+6 18+0-19+6 20+0-21+6 22+0-22+6\n84^ 102^ 107^ 37^\nGestation at PPROM (weeks)\n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   10 \nby gestation at birth; the rate for infants born under 28+0 weeks gestation was 53% and the \nrate for those infants born at or after 34+0 weeks gestation was 87% (Table A3). The \nunadjusted relative rate ratio of survival without severe morbidity was 1.23 (95% CI 1.04-\n1.47, p=0.027) per additional week of gestation at birth. \n \nSingleton pregnancy losses \nAmongst the 310 singleton pregnancies with known discharge status, a total of 256 (83%) \nresulted in pregnancy loss. Details of the type of pregnancy loss are given in Table 4.  \n \nThe rate of TFMR without expectant management reduced from 32% (25/79) when PPROM \noccurred at 16+0-17+6 weeks to 3% (1/33) with PPROM at 22+0-22+6 weeks gestation.  The \nrate of TFMR after expectant management was relatively similar at 8-18% across gestations \nstudied (Table 4). When PPROM occurred prior to 22+0 weeks the rate of birth or \nintrauterine death under 22+0 weeks (often called miscarriage) was relatively consistent at \n27-37%. There were more intrauterine deaths at 22+0 or more weeks gestation and neonatal \ndeaths when PPROM occurred between 22+0 and 22+6 weeks gestation (30% and 18% \nrespectively), compared to 9% and 8% respectively with PPROM prior to 22+0 weeks \ngestation (p<0.001 and p=0.05 respectively). This was largely because 39% of pregnancies \nended within a week of PPROM (Table 2) which meant that babies born after PPROM at 22+0 \nweeks gestation had a higher chance of birth with extreme prematurity. \n \n \n \nTable 4: Types of pregnancy loss in singleton pregnancies \n ^Pregnancies with missing discharge status are excluded: (5) after PPROM at 16+0-17+6 weeks; (3) after PPROM at 18+0-19+6 \nweeks; (8) after PPROM at 20+0-21+6 weeks; (4) after PPROM at 22+0-22+6 weeks \n \nMaternal outcomes \n \nAmongst the 268 women with singleton pregnancies who chose initial expectant \nmanagement 12% (33/268) developed maternal sepsis (Table 5). Three of these women \nbecame severely unwell; two died and a third was admitted to ITU and survived. These \nthree women also required surgical removal of the placenta. Two of the three women \ndeteriorated within 5 days of diagnosis of PPROM, and the third over a month after PPROM.  \n \nThe rate of maternal sepsis was not assessed amongst women that had TFMR without \nexpectant management (n=62), however there were no deaths or ITU admissions among \nthese women. The rate of maternal death with singleton pregnancies is therefore 2/330, \n0.61%, 95%CI 0.17-2.2%. \n \nTwenty two percent (51/227) of women with singleton pregnancies and expectant \nmanagement required surgery for placental removal. The rate of surgery for placental tissue \nNumber\nBirth or intrauterine death under 22+0 weeks gestation (miscarriage) 90 29% 26 33% 37 37% 27 27%\nIntrauterine death 22+0 or more weeks gestation 35 11% 3 4% 6 6% 16 16% 10 30%\nNeonatal death 28 9% 4 5% 8 8% 10 10% 6 18%\nTermination For Medical Reasons without expectant management 62 20% 25 32% 19 19% 17 17% 1 3%\nTermination For Medical Reasons after expectant management 41 13% 14 18% 13 13% 8 8% 6 18%\n310^ 79^ 99^ 99^ 33^\nSingleton pregnancies Whole cohort Gestation at PPROM (weeks)\n16+0-17+6 18+0-19+6 20+0-21+6 22+0-22+6\n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   11 \nremoval was 14% (14/103) amongst women with singleton pregnancies who had a TFMR. \nThis difference did not reach statistical significance (p=0.06). \n \n \n \nTable 5: Maternal complications of singleton pregnancies. Data for whole cohort presented as n, % of total and (95% \nconfidence interval). Data by gestational age at PPROM presented as n and % of total. p values compare outcomes by \ngestational age category at PPROM and are calculated by chi squared test. \n*Whether, or not, the mother developed sepsis was only asked for women who initially opted for expectant management. \nTherefore, these are expressed as a proportion excluding those with immediate TFMR. Women that had TFMR after \nexpectant management are included, therefore n=268. Surgery for placental removal is expressed as a proportion of all \nwomen.  \n \nMultiple pregnancies \n \nIn our cohort there were 38 women with multiple pregnancies (10%, 38/368), which is an \nover-representation as fewer than 2% of births nationally are from multiple pregnancies.[9] \nTwenty three were dichorionic-diamniotic twins (DCDA), 10 were monochronic-diamniotic \ntwins (MCDA), one was trichorionic triplets. Chorionicity was not determined in four twin \npregnancies.   \n \nIn six out of 30 twin pregnancies with expectant management, both infants survived to \nhospital discharge (20%). In a further five twin pregnancies (17%) one baby survived to \ndischarge from hospital. (Table 6). If one assumes that all infants with TFMR may have been \nliveborn, and those with missing information about hospital discharge did survive, then the \nsurvival to hospital discharge rate could be as high as 27% for both infants and as high as \n49% for a single twin infant at discharge. In the worst-case scenario with no livebirths in the \ncohort of pregnancies with TFMR and all those with missing discharge status having died \nthen the survival to hospital discharge rate would be 16% for both infants with an additional \n14% of pregnancies having a liveborn single infant surviving to discharge. The majority of \ntwin survivors were from DCDA pregnancies (Table 6). Six twin pregnancies (6/37, 16%) had \ndeath of a single infant (either intrauterine death or spontaneous birth under 22+0 weeks \ngestation) and TFMR of a second twin.  \n \nAmongst the 10 women with MCDA pregnancies there was only one pregnancy with survival \nof both infants, and a further two pregnancies with survival of a single infant. Importantly, \nsix out of 10 MCDA pregnancies had either laser coagulation or amnioreduction for twin-to-\ntwin transfusion syndrome prior to PPROM.  \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   12 \n \nTable 6: Infant outcomes for twin pregnancies. \n^ There were four twin pregnancies with undocumented chorionicity, these are included within the ‘All twin pregnancies’ \ndata only \n^^Missing data by chorionicity: DCDA discharge status for both infants (1); DCDA discharge status for a single infant (1); \nMCDA discharge status for a single infant (1); Unknown chorionicity discharge status for a single infant (1).  Livebirth status \nwas available for all twin infants. \n*Excludes pregnancies with termination for medical reasons (TFMR) of both infants \n** Excludes pregnancies with TFMR of both infants and discharge status missing for either or both infants \n \nThirty women with twin pregnancies (16 DCDA, 8 MCDA and 4 with chorionicity not \ndetermined) had expectant management but three (1 DCDA, 1 MCDA and 1 with \nchorionicity not determined) did not have complete information on infant morbidity. In the \ncohort with complete data, 17 out of 54 infants (31%) survived until discharge from hospital \n(6 sibling pairs and 5 single twins). Four of the 17 surviving infants (24%), all from DCDA \npregnancies, had severe morbidity when discharged.  \n \nThe rate of survival to hospital discharge without severe morbidity after expectant \nmanagement of twin pregnancies was therefore 13/54 (24%) of infants (4 sibling pairs and 5 \nsingle twins).  \n \nMaternal morbidity in twin pregnancies \n \nOverall, maternal morbidity in twin pregnancies was similar to that in singleton pregnancies \n(Tables 5 and 7). However, maternal sepsis was somewhat higher in women with twin \npregnancies that had expectant management or TFMR after expectant management (26% \ncompared to 12% for singletons; p=0.03). Two women with multiple pregnancies required \nintensive care treatment. \n \n  \nTable 7: Maternal outcomes of twin pregnancies. \n^ There were four twin pregnancies with undocumented chorionicity, these are included within the ‘All twin pregnancies’ \ndata only \nTwin pregnancies\nTotal number of pregnancies\n10/21 48% 3/9 33% 14/34 41%\n(43 -52%) (30 -40%) (38 -46%)\n1/21 5% 2/9 22% 3/34 9%\n(4 -35%) (20 -40%) (8 -32%)\n5/19 26% 1/8 13% 6/30 20%\n(22 -35%) (10 -20%) (16 -27%)\n3/19 16% 2/8 25% 5/30 17%\n(13 -52%) (20 -50%) (14 -49%)\n2 9% 1 10% 3 8%\n5 22% 1 10% 6 16%\nSurvival to hospital discharge of a single infant after expectant management**\nRange based on all twins\nTermination For Medical Reasons (TFMR) of both infants\nTermination For Medical Reasons (TFMR) of single infant only\nLivebirth of both infants after expectant management*\nRange based on all twins\nLivebirth of a single infant after expectant management* \nRange based on all twins\nSurvival to hospital discharge of both infants after expectant management**\nRange based on all twins\nChorionicity All twin \npregnancies^DCDA MCDA\n23^^ 10^^ 37^^\n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   13 \n*Whether, or not, the mother developed sepsis was only asked for women that initially opted for expectant management. \nTherefore, these are expressed as a proportion excluding those with immediate TFMR (n=35 for all women with twin \npregnancies). Women who had TFMR after expectant management are included. The remaining outcomes are expressed \nas a proportion of all women.  \n \n \nDiscussion \n \nStatement of principal findings  \n \nThe results of this national, population-based study of PPROM prior to 23 weeks gestation \nstarkly illustrate the diverse infant and maternal outcomes possible with this condition. \nWhilst 26% (54/207) of women with expectantly managed singleton pregnancies had an \ninfant that survived to hospital discharge, only 18% of women (38/207) had an infant that \nsurvived without severe morbidity. Maternal sepsis developed in 12% (33/268) of women \nwith singleton pregnancies and two women died. There is additional complexity because of \nthe uncertainty concerning infant outcomes inherent in 31% of women having TFMR.  \n \nComparison with previous studies \n \nThe two maternal deaths equate to a rate of 606 per 100 000 maternities with PPROM 16+0-\n22+6 weeks gestation, 95% CI 166-2183 per 100 000. This strikingly higher than the baseline \nUK maternal mortality rate of 11 per 100 000 maternities[12] and requires contextualising \nwithin the wider scientific literature.  \n \nIn the past decade over 700 women have been included in observational cohorts of PPROM \nat a similar gestations with no maternal deaths reported.[3,13–18] This concurs with no \nmaternal deaths reported in the most recent review on the topic,[19] and only a single \nmaternal death reported in the largest review, published in 2009, citing a publication from \n1988.[2,20] Therefore the maternal deaths are likely to be surprising to many clinicians. The \nkey difference between the current work and previous studies is that our cohort was \npopulation based as we surveyed all 194 consultant-led maternity units in the UK in contrast \nto previously published studies from five or fewer (often specialised) centres. In accordance \nwith the population-based approach second trimester PPROM associated with maternal \nsepsis and death has featured in three of the UK’s maternal mortality reports in the past \ndecade. [25–27] Population level data has also identified seven women who died after \nPPROM between 14+0 and 24+6 weeks gestation from 2001 to 2015 in France, giving an \nestimated chance of death of 45 per 100 000 maternities with this complication.[21] These \nwomen had similar causes of death to the current cohort; six were attributed to sepsis and \none to haemorrhage secondary to placenta accreta spectrum. Therefore, whilst the \nabsolute risk of maternal death with very early PPROM is likely to be within the lower range \nof our confidence intervals the population-based literature suggests that these are not \nisolated incidents, and may have previously been under-recognised because the literature \nhas been based on data from a small number of centres. \n \nThe rate of maternal infectious morbidity in previous studies of expectantly managed \nsingleton pregnancies is similar to our 12%.[3,13,22]  In this cohort maternal sepsis was \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   14 \nhigher in twin pregnancies, also in concordance with previous work.[23]  Surgery for \nplacental removal is a less well recognised complication of birth after PPROM but occurred \nin 20% of our singleton cohort, including the three women who became severely unwell. \nThis rate concurs with a cohort of midtrimester PPROM cases in Ireland.[3] The combination \nof requiring surgery for placental removal and sepsis should alert clinicians to the possibility \nof maternal deterioration.  \n \nThe infant survival to hospital discharge rate of 26% for expectantly managed singleton \npregnancies, and the range of possible survival of 16-54% when pregnancies with TFMR and \nunknown outcomes are accounted for are broadly in keeping with observational studies \nover the past decade, that reported rates of 17-40% with similar gestations of \nPPROM.[13,15,24–28] Amongst the infants that survived to hospital discharge, 70% avoided \nsevere morbidity by our definition, this is also in keeping with recently published \nobservational studies.[13,22,26,27] \n \nThe severe neonatal morbidity definition was chosen for consistency with Kibel et al (2016) \nwho related neonatal outcomes at hospital discharge following pregnancies complicated by \nPPROM at 20-24 weeks gestation to outcomes at a corrected age of 18 – 21 months. Of 24 \nneonates with grade 3 or 4 IVH and/or requirement for oxygen at 36 weeks postmenstrual \nage and/or grade 3 or more retinopathy of prematurity, 8 (30%) had moderate-severe \nmorbidity when they were 18-21 months old. Among 27 babies who were born following \npregnancies complicated by PPROM at 20-24 weeks gestation but who did not have any IVH, \nROP or oxygen requirement outcomes at 36 weeks, none had moderate-severe morbidity \nwhen they were 18-21 months old.[22]. Our results are likely to be indicative of long-term \nmoderate-severe morbidity in a similar proportion of children. \n \nWithin this study women that had expectant management after PPROM at 16+0-17+6 weeks \ngestation and gave birth after 23 weeks gestation had comparable livebirth and infant \nmorbidity rates to those who experienced PPROM at 22+0-22+6 weeks gestation. This could \nbe due to lack of statistical power because only seven infants survived to discharge after \nPPROM at 16+0- 17+6 weeks gestation. It is also possible that when PPROM occurred earlier \nin pregnancy more pregnancies with less favourable characteristics had a TFMR, negating \nany negative effect of gestation of PPROM on infant morbidity.  However a study from The \nNetherlands, with a rate of termination of only 2%, also found no difference in infant \nmorbidity after at PPROM at  ≥13- <20 weeks (n=21) compared to ≥20- <24 weeks \n(n=41).[26] We therefore suggest that the impact of gestation when PPROM occurs on \ninfant morbidity needs further evaluation. \n \nThe median gestational age at birth of surviving infants was 29+4 weeks gestation, and the \nmedian length of hospital stay after birth for surviving infants was 59 days, IQR 17-100 days. \nWhilst this is shorter than contemporary studies from Australia [15] (median 76 days, IQR \n44-111 days) and Japan [29] (mean 155 days, standard deviation (SD) 53 days), it is still a \nsignificant amount of time and likely to have a substantial impact upon the whole family in \nthe medium term. We suggest this information should be included in patient counselling. \n \nAs gestational age at birth advanced the chance of livebirth and infant survival to hospital \ndischarge improved, as expected.[2] However, even among those born after 34 weeks \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   15 \ngestation two infants (2/15, 13%) had severe morbidity, illustrating the complexities of \nthese cases and the need for ongoing multidisciplinary team care, including neonatologists, \neven at relatively advanced gestations. \n \nInfants of DCDA multiple pregnancies with PPROM appeared to have comparable pregnancy \noutcomes to singletons. This is in keeping with previous literature.[30] MCDA pregnancies \nhad lower infant survival, but 60% of these pregnancies also had pathologies of \nmonochorionicity such as twin to twin transfusion syndrome and selective growth \nrestriction. Therefore, the pathologies unique to monochorionic pregnancies are likely to be \nthe key contributor to the mortality in these instances. \n \n \nStrengths and Limitations \n \nThe UKOSS infrastructure has enabled the largest population-based study of PPROM prior to \n23 weeks’ gestation. There is, however, unavoidable uncertainty concerning infant \noutcomes due to 31% of our population opting for TFMR and inability to follow up all \ninfants, particularly those that moved hospital as part of their care.  \n \nThe 0.04% incidence of PPROM between 16+0-22+6 weeks may be an under estimation due \nto the possibility of under-reporting of cases, particularly during the covid-19 pandemic. \nHowever, we have no evidence of biased reporting which might influence the \ngeneralisability of the results. A recent French analysis of hospital episode statistics for the \nwhole nation found a prevalence of 0.2% of PPROM 14+0-24+6 weeks, and the authors \nconsidered this figure to be likely an under-estimation, noting that it is impossible with \ncoded data to fully validate the diagnosis.[21] The number of cases reported was lowest \nbetween July 2020 and December 2020, possibly because of staffing pressures, or \npotentially due to a lower rate of PPROM secondary to public health control measures such \nas lockdowns. \n \nOur survey did not include questions about whether mothers who opted straight for \ntermination of pregnancy after PPROM developed sepsis, and to simplify case reporting and \ncapture the maximum number of cases the inclusion criteria were kept brief and questions \nregarding fetal anomalies, whether identified at the time of PPROM or not, were not \nincluded. The survey reporters were not asked whether women that opted for a TFMR, or \nhad a baby loss, also had a fetus with a life-threatening anomaly. As the presence of such a \nfetal anomaly may have been a possibility in some instances, their presence may have \ninfluenced the decision to terminate the pregnancy. A limited range of neonatal outcomes, \nand no outcomes after hospital discharge, were collected because of resource constraints.   \n \n \nImplications for clinical practice \n \nThis study illustrates that women with PPROM prior to 23 weeks gestation have some of the \nhighest risks of infant and maternal morbidity that a clinician will face amongst the women \nthey care for. An individual obstetrician is likely to only see one or two women with \nextremely preterm PROM a year, thereby being unable to build a wealth of experience \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   16 \nrelevant to the condition. In accordance with other recent studies we also highlight that \nwomen opting for TFMR do not avoid all severe morbidity.[23,31] \n \nIn current UK practice pregnant women prior to 20 weeks gestation are often cared for on \ngynaecology rather than obstetric wards. The infant mortality that occurred was largely \nattributable to birth prior to viability and extreme prematurity because 39% of women gave \nbirth in the week after PPROM, and a further 21% gave birth in the following week. This \npattern was consistent across the gestational ages of PPROM studied, and is consistent with \nprevious work. [13,26] We suggest that pregnancies with PPROM under 23 weeks gestation \nhave a high likelihood of needing support from infant bereavement teams, neonatal teams \nand/or maternal critical care. Obstetricians with an interest in this condition, alongside \ndedicated midwives, are likely to be best placed to co-ordinate these elements of care in the \nimmediate aftermath of PPROM and for the remainder of the pregnancy. Neonatal teams \ncan note that about half of babies who were admitted to neonatal units following PPROM \nbetween 16+0-22+6 weeks survived. A majority of survivors did not have the morbidities we \ncaptured. Among survivors, the likelihood of the morbidities we captured was lower with \neach week of completed pregnancy. \n \nWe suggest that teams with expertise in management of very early PPROM, along with \npatient representatives, work together to develop guidelines for care.  \n \nImplications for research \nThis study provides baseline data on infant and maternal outcomes of pregnancies with \nPPROM under 23 weeks gestation upon which studies aiming to improve outcomes can be \nplanned. Interventions may be novel treatments aimed at treating the pathology, or care \nbundles, including training, to optimise care and delivery timing using interventions already \navailable.  \n \nMaterials for women and families facing very early PPROM will need to consider the \ninherent uncertainty within the data secondary to 31% of women with singleton \npregnancies opting for TFMR. The optimal way to communicate such wide uncertainties \nwithin the data to a wider audience, and support families with such complex pregnancies, is \nyet to be determined. \n \nThe current study is only able to comment on severe morbidity in infants at discharge from \nhospital, and previous work suggests that 70% of these infants will not have significant \nmorbidity when aged 18-21 months.[22] The rate of long term disability in offspring is the \nmost desired information for prospective parents. Future research needs to incorporate \nways of capturing this information. \n \nThere is likely to be substantial unmeasured psychological morbidity related to this \ncondition. The optimal way of managing very early PPROM to support psychological \nwellbeing of families requires further consideration. \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   17 \nConclusions \n \nWomen with PPROM prior to 23 weeks gestation can have favourable maternal and \npregnancy outcomes. However, a significant proportion of these pregnancies are \ncomplicated by maternal morbidity and infant mortality and morbidity.  All clinicians who \ncare for these families need to be conscious of the risk of maternal sepsis and death. The \ndata presented will be helpful in counselling families facing PPROM prior to 23 weeks \ngestation and should be incorporated into updates of clinical guidelines. \n \n \nAcknowledgements \nWe would like to acknowledge the assistance of UK Obstetric Surveillance System reporting \nclinicians, the Little Heartbeats volunteers and the UK Obstetric Surveillance System \nSteering Committee without whose support this research would not have been possible. We \nwould also like to thank Mrs Tracey Ricketts for administrative assistance. \n \nData sharing \nData cannot be shared publicly because of confidentiality issues and potentially identifiable \nsensitive data as identified within the Research Ethics Committee application/approval. \nRequests to access the data can be made by contacting the National Perinatal Epidemiology \nUnit data access committee via general@npeu.ox.ac.uk. \n \nRole of the funding source \nThis work was supported by Wellbeing of Women in partnership with Little Heartbeats \n(Award Ref RG2241). The views expressed are those of the authors and not necessarily \nthose of Wellbeing of Women or Little Heartbeats. CC, the founder of Little Heartbeats, \nprovided PPI input into the study but ultimate responsibility for the final study design; the \ncollection, analysis, and interpretation of data; the writing of the report; and the decision to \nsubmit the paper for publication was held by the academic researchers, LG, MK and ZA. All \nauthors had full access to all the data in the study and can take responsibility for the \nintegrity of the data and the accuracy of the data analysis. MK is an NIHR Senior Investigator \n(NIHR201333). The views expressed are those of the authors and not necessarily those of \nthe NHS, the NIHR or the Department of Health and Social Care. \n \n \nTransparency statement \nLG (the manuscript's guarantor) affirms that the manuscript is an honest, accurate, and \ntransparent account of the study being reported; that no important aspects of the study \nhave been omitted; and that any discrepancies from the study as originally planned have \nbeen explained. \n \nCompeting interests \nAll authors have completed the ICMJE uniform disclosure form at \nwww.icmje.org/disclosure-of-interest/ and declare: All authors received a grant from \nWellbeing of Women in partnership with Little Heartbeats (Award Ref RG2241) in relation to \nthe submitted work. MK also holds National Institute for Health Research (NIHR) and \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   18 \nHealthcare Quality Improvement Partnership grants, and ZA also has an NIHR Infrastructure \ngrant covering Cochrane Pregnancy Childbirth. CC is the founder, a volunteer and president \nof Little Heartbeats and to support Little Heartbeats she has received donations from the \npublic and royalties for a song that she co-wrote in memory of her daughter, Sinead. LG, AC, \nMT and DR declare no further competing interests. \n \n \n \n \n \n \n \nReferences \n \n1  Thomson AJ. Care of Women Presenting with Suspected Preterm Prelabour Rupture \nof Membranes from 24+0 Weeks of Gestation: Green-top Guideline No. 73. BJOG An \nInt J Obstet Gynaecol 2019;126:e152–66. doi:10.1111/1471-0528.15803 \n2  Waters TP, Mercer BM. The management of preterm premature rupture of the \nmembranes near the limit of fetal viability. Am J Obstet Gynecol 2009;201:230–40. \ndoi:10.1016/j.ajog.2009.06.049 \n3  Linehan LA, Walsh J, Morris A, et al. Neonatal and maternal outcomes following \nmidtrimester preterm premature rupture of the membranes: a retrospective cohort \nstudy. BMC Pregnancy Childbirth 2016;16:25. doi:10.1186/s12884-016-0813-3 \n4  Little Heartbeats. What to expect- in hospital. 2023.https://www.little-\nheartbeats.org.uk/what-to-expect-c10zs (accessed 9 Jan 2023). \n5  Knight M, Kurinczuk JJ, Tuffnell D, et al. The UK Obstetric Surveillance System for rare \ndisorders of pregnancy. BJOG 2005;112:263–5. doi:10.1111/j.1471-\n0528.2005.00609.x \n6  UKOSS. UKOSS Extremely Preterm Prelabour Rupture of the Membranes Study. \n2019.https://www.npeu.ox.ac.uk/assets/downloads/ukoss/forms/UKOSS EPPROM \nv1.0 - 21-Oct-2019.pdf (accessed 16 Jan 2023). \n7  Northern Ireland Statistics and Research Agency. Monthly Births. \n2022.https://www.nisra.gov.uk/publications/monthly-births (accessed 9 Jan 2023). \n8  General Register Office for Scotland. Vital Events Reference Tables 2020. \n2021.https://www.nrscotland.gov.uk/statistics-and-data/statistics/statistics-by-\ntheme/vital-events/general-publications/vital-events-reference-tables/2020/list-of-\ndata-tables#section3 (accessed 9 Jan 2023). \n9  Office for National Statistics. Births in England and Wales 2020. \n2022.https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmar\nriages/livebirths/datasets/birthcharacteristicsinenglandandwales (accessed 9 Jan \n2023). \n10  Von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of \nObservational Studies in Epidemiology (STROBE) Statement: Guidelines for reporting \nobservational studies. Bull World Health Organ 2007;85:867–72. \ndoi:10.2471/BLT.07.045120 \n11  MacTier H, Bates SE, Johnston T, et al. Perinatal management of extreme preterm \nbirth before 27 weeks of gestation: A framework for practice. Arch Dis Child Fetal \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   19 \nNeonatal Ed 2020;105:F232–9. doi:10.1136/archdischild-2019-318402 \n12  Knight M, Bunch K, Tuffnell D, et al. Saving Lives, Improving Mothers’ Care - Lessons \nlearned to inform maternity care from the UK and Ireland Confidential Enquiries into \nMaternal Deaths and Morbidity 2017-19. Oxford: 2021.  \n13  Pylypjuk C, Majeau L. Perinatal outcomes and influence of amniotic fluid volume \nfollowing previable, preterm prelabor rupture of membranes (Ppprom): A historical \ncohort study. Int J Womens Health 2021;13:627–37. doi:10.2147/IJWH.S303120 \n14  Roberts D, Vause S, Martin W, et al. Amnioinfusion in preterm premature rupture of \nmembranes (AMIPROM): a randomised controlled trial of amnioinfusion versus \nexpectant management in very early preterm premature rupture of membranes--a \npilot study. Health Technol Assess 2014;18:1–135. doi:10.3310/hta18210 \n15  Sim WH, Ng H, Sheehan P. Maternal and neonatal outcomes following expectant \nmanagement of preterm prelabor rupture of membranes before viability. J Matern \nNeonatal Med 2020;33:533–41. doi:10.1080/14767058.2018.1495706 \n16  van der Marel I, de Jonge R, Duvekot J, et al. Maternal and Neonatal Outcomes of \nPreterm Premature Rupture of Membranes before Viability TT  - Maternale und \nneonatale Mortalität und Morbidität bei vorzeitigem Blasensprung vor Beginn der \nLebensfähigkeit des Kindes. Klin Padiatr 2016;228:69–76. \n17  Panzer A, Dotters-Katz S, Smid M, et al. Factors Associated with Previable Delivery \nfollowing Second Trimester Rupture of Membranes. Am J Perinatol 2019;36:812–7. \ndoi:10.1055/s-0038-1675373 \n18  Miyazaki K, Furuhashi M, Yoshida K, et al. Aggressive intervention of previable \npreterm premature rupture of membranes. Acta Obstet Gynecol Scand 2012;91:923–\n9. doi:10.1111/j.1600-0412.2012.01432.x \n19  Sim WH, Araujo Júnior E, Da Silva Costa F, et al. Maternal and neonatal outcomes \nfollowing expectant management of preterm prelabour rupture of membranes \nbefore viability. J Perinat Med 2017;45:29–44. \n20  Moretti M, Sibai BM. Maternal and perinatal outcome of expectant management of \npremature rupture of  membranes in the midtrimester. Am J Obstet Gynecol \n1988;159:390–6. doi:10.1016/s0002-9378(88)80092-9 \n21  Abrahami Y, Saucedo M, Rigouzzo A, et al. Maternal mortality in women with pre-\nviable premature rupture of membranes: An analysis from the French confidential \nenquiry into maternal deaths. Acta Obstet Gynecol Scand 2022;:1395–402. \ndoi:10.1111/aogs.14452 \n22  Kibel M, Asztalos E, Barrett J, et al. Outcomes of Pregnancies Complicated by Preterm \nPremature Rupture of Membranes Between 20 and 24 Weeks of Gestation. Obstet \nGynecol 2016;128:313–20. doi:10.1097/AOG.0000000000001530 \n23  Dotters-Katz SK, Heine RP, Grotegut CA. Medical and infectious complications \nassociated with pyelonephritis among pregnant women at delivery. In: Infectious \nDiseases in Obstetrics and Gynecology. Hindawi Limited 2013. 124102. \ndoi:10.1155/2013/124102 \n24  Hunter TJ, Byrnes MJ, Nathan E, et al. Factors influencing survival in pre-viable \npreterm premature rupture of  membranes. J Matern neonatal Med  Off J  Eur Assoc \nPerinat Med Fed Asia Ocean Perinat Soc Int Soc Perinat Obstet 2012;25:1755–61. \ndoi:10.3109/14767058.2012.663824 \n25  van der Marel I, de Jonge R, Duvekot J, et al. Maternal and Neonatal Outcomes of \nPreterm Premature Rupture of Membranes before  Viability. Klin Padiatr \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint \n\n   20 \n2016;228:69–76. doi:10.1055/s-0041-111174 \n26  Van Der Heyden JL, Van Der Ham DP, Van Kuijk S, et al. Outcome of pregnancies with \npreterm prelabor rupture of membranes before 27 weeks’ gestation: A retrospective \ncohort study. Eur J Obstet Gynecol Reprod Biol 2013;170:125–30. \ndoi:10.1016/j.ejogrb.2013.06.012 \n27  Lorthe E, Torchin H, Delorme P, et al. Preterm premature rupture of membranes at \n22-25 weeks’ gestation: perinatal and 2-year outcomes within a national population-\nbased study (EPIPAGE-2). Am J Obstet Gynecol 2018;219:298.e1-298.e14. \ndoi:10.1016/j.ajog.2018.05.029 \n28  Azria E, Anselem O, Schmitz T, et al. Comparison of perinatal outcome after pre-viable \npreterm prelabour rupture of membranes in two centres with different rates of \ntermination of pregnancy. BJOG 2012;119:449–57. doi:10.1111/j.1471-\n0528.2011.03265.x \n29  Sorano S, Fukuoka M, Kawakami K, et al. Prognosis of preterm premature rupture of \nmembranes between 20 and 24 weeks of gestation: A retrospective cohort study. Eur \nJ Obstet Gynecol Reprod Biol X 2020;5:100102. doi:10.1016/j.eurox.2019.100102 \n30  Wong LF, Holmgren CM, Silver RM, et al. Outcomes of expectantly managed \npregnancies with multiple gestations and preterm premature rupture of membranes \nprior to 26 weeks. Am J Obstet Gynecol 2015;212:215.e1-215.e9. \ndoi:10.1016/j.ajog.2014.09.005 \n31  Sklar A, Sheeder J, Davis AR, et al. Maternal morbidity after preterm premature \nrupture of membranes at <24 weeks’ gestation. Am J Obstet Gynecol \n2022;226:558.e1-558.e11. doi:10.1016/j.ajog.2021.10.036 \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprintthis version posted March 9, 2023. ; https://doi.org/10.1101/2023.03.07.23286863doi: medRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}