{"paper_id":"a3ce1bee-2c95-4c5a-8284-d23e52560f63","body_text":"The prevalence of bleeding disorders in women with proven endometriosis: results \nof a screening study. \nSarah Heynemann1, Emma Johnson1, Melissa Cameron1, Ray Dauer2, Elya Moore3, Chris Barnes4, Sonia R \nGrover1 \n1Dept Obstetric and Gynaecology, University of Melbourne, Mercy Hospital for Women; 2Department of \nHaematology, Austin Health;3Department of Microbiology, University of Melbourne; 4Department of Haematology \nRoyal Children’s Hospital \nCorresponding author: Sonia Grover, Department of Obstetrics and Gynaecology, Mercy Hospital for Women, \nHeidelberg 3084, Victoria Australia  Tel+61 3 93455800; Fax +61 3 93456668   s.grover@bigpond.net.au \nFunding sources:  University of Melbourne, Advanced Medical Science program; Australian Gynaecology Endoscopy \nand Surgery Society Ltd; Siemens: provision of the PFA-100 Analyzer machine for this project \n \nAbstract \nObjective: To establish the rate of bleeding disorders amongst women with confirmed endometriosis and \nclarify predictive features on history. \nMethods: Eight-two women with confirmed diagnosis of endometriosis were recruited from a tertiary \nwomens’ hospital. General bleeding tendency was evaluated with a modified internationally standardised \nquestionnaire, and a bleeding score was calculated. Menstrual loss was evaluated by obtaining current and \nheaviest-ever menses profile using Pictorial Blood loss Assessment Chart(PBAC).  Forty-one women \nsatisfied criteria for a significant bleeding tendency and underwent coagulation tests including Prothrombin \ntime(PT), activated partial thromboplastin time(aPTT), Platelet Function Analyser-100(PFA-100) and von \nWillebrand factor(vWF)/Factor VIII(FVIII) studies.  \nResults :  The prevalence of abnormal screening haemostasis tests for the population was 17.72% (95% CI \n10.04-27.94) for all tests, 8.54%(95% CI 3.50-16.80) for vWF tests, and 12.99%( 95% CI 6.41-22.59) for \nPFA-100. Receiver Operator Characteristic curve analysis demonstrated a ‘very good’ performance of the \nbleeding score as a diagnostic test for haemostatic abnormalities for nulliparous women with \nendometriosis. Bleeding symptoms individually were not predictive. Logistic regression suggested the \ncombination of mittleschmerz, cutaneous bleeding symptoms, heavy menstrual bleeding and prolonged \nbleeding from minor wounds as predictive for haemostatic abnormalities (p=0.0197).  \nConclusion: The prevalence of abnormal bleeding tests in women with endometriosis on preliminary \ntesting was higher than the population rate of bleeding abnormalities and thus people with histologically \nproven endometriosis warrant a higher index of suspicion with respect to testing for mild bleeding \ndisorders. \nKey words: endometriosis, mild bleeding disorders, vonWillebrand disease, platelet disorders \n \n Bleeding disorders in women with endometriosis \n \n \n \n \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: 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\nIntroduction \nEndometriosis is a benign condition primarily affecting women of reproductive age. A definitive diagnosis \nof endometriosis is based on histopathological confirmation of suspected endometriotic biopsies taken \nduring surgery and this remains the gold standard in clinical guidelines from many national and \ninternational societies (As-Saine, Black et al 2019). A systematic review of the correlation between \nhistology and surgical visualization reported a ‘negative’  laparoscopy is likely to be accurate in excluding \nendometriosis, however a ‘positive’ laparoscopy, at which disease is visualized surgically, is less \ninformative, hence histological confirmation is advocated(Wykes, Clark et al. 2004).  \nAlthough there is some preliminary work suggesting that a non-invasive diagnostic test may be on the \nhorizon (Bendifallah, Suisse et al 2022), estimates of endometriosis disease prevalence are limited to \nhospital or surgical audits. Whilst a population prevalence rate of 10% has been proposed (Eskenazi and \nWarner 1997), and is generally accepted,  inconsistency  in case definition for endometriosis continues to \nbe a significant concern (Koninckx, Ussia et al 2021) and rates vary depending on population selection with \nthe highest rates in those with infertility(25-50%) or chronic pelvic pain (71-87%) (As-Saine, Black et al \n2019).  Identifying the lack of symptoms or cluster of present symptoms that predict for a diagnosis of \nendometriosis remains a challenge (Surrey, Carter et al 2017) as does the lack of correlation between \nseverity of symptoms and severity of endometriosis (Agarwal, Chapron et al 2019). \nAlthough retrograde menstruation is considered a physiological process present in close to 100% of women \nundergoing laparoscopy at the time of menses (Halme, Hammond et al. 1984), the challenge is \nunderstanding why endometriosis does not occur in all women. Although ‘implantation’ is the prevailing \ntheory for the pathogenesis of peritoneal endometriosis, other factors including immune dysfunction, \noxidative stress and stem cells may all contribute (Burney and Guidice 2012).  Observations to support this \ninclude the higher prevalence of retrograde menstruation in baboons and women with spontaneous \nendometriosis than those with a normal pelvis and experimental implantation (D'Hooghe and Debrock \n2002). In studies on the risk factors for endometriosis, a  consistent finding is that exposure to \nmenstruation, with earlier age of menarche, shorter cycle duration (Missmer and Cramer 2003), and longer \nand heavier menses(Cramer, Wilson et al. 1986; Vercellini, De Giorgi et al. 1997) are all associated with an \nincreased risk of disease, whilst there is a decreased risk with increasing parity (Eskenazi and Warner \n1997). \nGiven the evidence that the volume of retrograde menstrual loss (D'Hooghe and Debrock 2002) appears to \nrelate to endometriosis risk, it may be more likely that women who are unable to clear this magnified \npotential stimulus for disease (Cramer and Missmer 2002) are at higher risk. Heavy and prolonged \nmenstrual bleeding are common symptoms amongst women with bleeding  disorders, and a higher \nprevalence of bleeding disorders amongst women with heavy menstrual bleeding (HMB) compared to the \ngeneral population has been reported, with a 13% rate for vonWillebrand Disease(vWD) (Shankar, Lee et \nal. 2004).  For women with vWD, menses may pose a considerable problem and impact on quality of life \n(Kirtava, Drews et al. 2003; Lee and Kadir 2005; James, Ragni et al. 2006); additionally,  bleeding may \noccur during pregnancy, postpartum, and at ovulation (Kouides, Phatak et al. 2000; James 2005).  Although \nthe presence of endometriosis was not histologically verified, in a survey of women with vWD, 30% of \nwomen reported having been told at some stage by a doctor that they had endometriosis (Kirtava, Drews et \nal. 2003). Despite the epidemiological evidence suggesting an affiliation between endometriosis and \nbleeding disorders (Vercellini, De Giorgi et al. 1997) and a report on a clinical observation( Mitri and \nCasper 2015), there have been no studies to date to confirm this. \nAlthough the reported rate of vWD in the general population has been 1% (Rodeghiero, Castaman et al. \n1987; Werner, Broxson et al. 1993), both of these epidemiological studies did not focus on medically \nsignificant or symptomatic bleeding, and with subsequent follow-up many of these individuals did not \ndevelop symptomatic or significant bleeding. A primary care study exploring the prevalence of \nsymptomatic vWD nevertheless concluded that the population rate was at least one in a thousand \n(Bowman, Hopman et al 2010).  \nPlatelet function disorders are thought to be of a similar frequency to vWD (Lusher 1999). Other bleeding \ndisorders are rare and have prevalence rates in the order of  1 in 10,000 for Factor IX (Peyvandi, Duga et al. \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\n2002), to 1 in 1,000,000 for Factor XI  (Mannucci and Tuddenham 2001; Peyvandi, Duga et al. 2002) , \nalthough an increased prevalence occurs in populations of Ashkenazi Jewish descent(Peyvandi, Duga et al. \n2002).   \nDespite bleeding disorders being recognized as a significant predisposing factor for HMB in \npremenopausal women, identification of these women has remained low. A survey in 2006 of obstetricians \nand gynaecologists revealed that only 2% of gynaecologists would consider investigating a 35 year-old \nwoman with HMB for a mild bleeding disorder (Chi, Shiltagh et al 2006). \nIdentifying women with a heightened bleeding tendency is challenging. Clinical determination of \n‘significant’ or ‘excessive’ bleeding tendency is fraught by non-specific symptoms and heterogeneity in \nclinical manifestations (Greaves and Watson 2007), with ‘haemorrhagic  symptoms’, real or perceived, \npresent in apparently normal individuals (Sadler 2003; Rodeghiero, Tosetto et al. 2007; Rodeghiero 2008). \nA structured bleeding history has been shown to have some utility (Sramek, Eikenboom et al. 1995; \nRodeghiero, Castaman et al. 2005; Tosetto, Rodeghiero et al. 2006; Philipp, Faiz et al. 2008) although the \ncriteria for concluding that there is a bleeding disorder influences the degree of predictiveness or utility. In \nthe widely accepted study by Rodegheiro, specificity of a bleeding history questionnaire was >90% and \nsensitivity 80% (Rodeghiero, Castaman et al. 2005). An additional tool that has been used for attempting to \nidentify women with heavy menses is the pictorial blood loss assessment chart (PBAC) (Higham, O'Brien \net al. 1990), with the combination (ie., questionnaire plus PBAC) used by some – either stepwise to identify \nwomen for further screening  (Kadir, Economides et al. 1998), or as part of a multi-component screening \ntool (Philipp, Faiz et al. 2008). \nThus this pilot study aimed to investigate the rate of bleeding disorders amongst women with proven \nendometriosis, and clarify predictive features on clinical history by utilizing both a standardised \nquestionnaire and the PBAC.  \nMethodology \nWomen aged 16-50 years with a history of histologically confirmed endometriosis were recruited from the \nendosurgery and gynaecology clinics of the Mercy Hospital for Women (MHW) and private rooms of the \nclinicians involved between October 2008 and March 2009. In the absence of histology, surgical \nvisualization and/or photographs of distinctive endometriomas with obliteration of the Pouch of Douglas \nwere considered sufficient evidence of endometriosis. Women with an intellectual disability or language \ndifficulties, or a previously diagnosed bleeding disorder were excluded. Women with adenomyosis were \nonly included if they also had proven endometriosis. \nFollowing recruitment and written consent, general and menstrual bleeding tendency were evaluated using \na standardized questionnaire and PBACs administered by SH or EJ. General bleeding tendency was \ninvestigated using a modified version of the standardized questionnaire originally proposed by Rodeghiero \net al. Each symptom could be considered ‘trivial’ or ‘significant’, and details were agreed upon via research \nteam consensus prior to data collection following specialist haematological and gynaecological input. \nSymptoms included epistaxis, cutaneous bleeding, oral cavity bleeding, gastrointestinal bleeding, \npostpartum haemorrhage, muscle haematomas, prolonged bleeding from minor wounds/after teeth \nextraction, and HMB; with 0-3 points allocated according to symptom severity. Modification to this \nquestionnaire included the addition of mittleschmerz (midcycle, ovulation related pain) (Kouides, Phatak et \nal. 2000; James 2005; Lee, Chi et al. 2006) and family history of excessive bleeding(Sramek, Eikenboom et \nal. 1995; Jayasinghe, Moore et al. 2005) given research subsequent to development of the original \nquestionnaire by Rodeghiero et al (2005) noting these to be discriminatory symptoms for bleeding \ndisorders.  A summative bleeding score was calculated based on questionnaire responses. Women were \nidentified as warranting haemostatic testing if they had a score >4 for nulliparous women, >5 for parous \nwomen, or the presence of \n≥  3 symptoms (“presence’ was defined by achievement of ≥ 1 point, or ≥ 2 points \nfor family history). Information regarding current medications, history of other gynaecological conditions, \nanaemia and liver disease was also recorded. Modification to the bleeding score was undertaken to promote \nhigh sensitivity of the tool to maximise likelihood of identifying women with clinical history suspicious for \na bleeding disorder for subsequent haemostatic testing, with lowering of the score for women who had not \nexperienced the haemostatic challenge of childbirth. The PBAC was administered to evaluate both current \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\n(if applicable) and ‘heaviest ever’ menstrual profile. A score of ≥ 100points for the PBAC was considered \nevidence of HMB as has been used in the original study by Higham et al (1990). \nWomen who satisfied the described criteria underwent testing for possible vWD and platelet dysfunction. \nBlood was taken by experienced phlebotomists and all test processing except for the PFA-100® (Platelet \nFunction Analyser-100) performed by Austin Hospital Pathology. Where possible the day of the woman’s \nmenstrual cycle was recorded.  Haemostatic evaluation included prothrombin time (PT) (using Hemosil \nRecombiPlasTin ®.  Instrumentation Laboratory, Lexington, Ma, USA); Activated Partial Thromboplastin \nTime (APTT) (performed using Triniclot HS, which uses purified phospholipids containing micronised \nsilica as the activator, Trinity Biotech, USA); von Willebrand Factor Antigen (vWF:Ag) (performed using \nthe HemosIL\nTM Von Willebrand Factor Antigen kit.  Instrumentation Laboratory, Lexington, Ma, USA); \nvWF Ristocetin Cofactor (vWF:RCo) (performed using stabilized platelets and ristocetin in a lyophilized \nform, from Dade Behring/Siemens, Inc Newark, DE, USA); vWF:collagen binding assay (vWF:CBA) \n(performed using an ELISA based method from Life Therapeutics.  Life Therapeutics, Frenchs Forest, \nNSW, Australia); Factor VIII:C. (Factor VIII activity was performed using a one stage APTT based \nmethod.  Factor VIII deficient plasma was obtained from Dade Behring/Siemens, Inc Newark, DE, USA). \nPlatelet function was assessed using PFA-100® (Dade Behring/Siemens, Inc Newark, DE, USA), utilising \nboth Collagen Epinephrine (C-EPI) and Collagen Adenosine Diphosphate (C-ADP) cartridges  using 3.2% \ncitrated whole blood between 30 minutes and 4 hours after blood draw as recommended (Hayward, \nHarrison et al. 2006).  \nResults for vWF:RCo were considered abnormal if <50%; vWF:Ag and vWF:CB results between 50-55% \nwere considered borderline and in the context of this screening study, abnormal.  The haematocrit value \nhad to be between 0.35 – 0.50, and platelets between 150-500 x 10\n9/L for specimens to be included for \nPFA-100 ® testing. Reference ranges for PFA-100 ® C-EPI Closure Time (CT)  ≤  170 seconds,  C-ADP \nCT ≤ 115secs. Where possible, testing was rescheduled if patients reported having used a list of common \nnonsteroidal anti-inflammatory drugs (NSAIDs) or aspirin in the 2 weeks prior to PFA-100 ®, otherwise \nthe number of days since last NSAID usage was recorded. \n \nWomen who returned an ‘abnormal’ result for any of the initial tests performed, according to stated \nreference ranges, were included for analysis in this study. For the purpose of prevalence estimates, women \nwho did not satisfy criteria for increased bleeding tendency and were not tested were assumed to have \nnormal haemostasis.\n \n \nEndometriosis \nClinical notes and photographs available from the operation/s at which endometriosis was diagnosed were \nreviewed by a single gynaecologist with an interest in endometriosis. \n \nStatistical analysis \nQuestionnaire and PBAC data were entered into an Excel database, blinded to blood test results, and \nanalysis was performed using STATA™ 8.2. P<0.05 was considered statistically significant. \nPrevalence rates of haemostatic abnormalities were calculated separately for vWF/FVIII, PFA-100® and \nall tests combined. Spearman Rank Correlation was performed to determine if non-parametric continuous \nvariables such as bleeding score, number of symptoms, PBAC score and age were independent of each \nother. Mann-Whitney tests were performed to examine the effect of parity status on distribution of bleeding \nscore, number of symptoms and PBAC scores; and to determine if PBAC scores differed between women \nwith ‘abnormal’ and ‘normal’ haemostasis results. Wilcoxon signed rank test was performed to determine if \n‘current’ and ‘heaviest ever’ PBAC scores differed significantly. Receiver Operator Characteristic (ROC) \ncurve analysis was performed to determine the individual value of symptoms, bleeding score, number of \nsymptoms, and PBAC as diagnostic tests for initial haemostatic abnormalities. ROC curves display the \ncompromise between sensitivity and specificity at different cut-offs of a diagnostic test (McPherson and \nPincus 2007). Area under the curve (AUC) can be calculated as a summary measure, with <0.7 considered \n‘poor’ and >0.9 considered ‘excellent’ (McPherson and Pincus 2007).  Step-wise backward selection \ngrouped linear logistic regression was performed to determine the most predictive combination of \nsymptoms. \n \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\nEthical approval \nEthical approval was granted by MHW Human Research and Ethics Committee for the project (R08/15).  \n \nResults \nBetween October 2008 and March 2009, 105 women with endometriosis were recruited. Of these 23 were \nsubsequently excluded from the study with 9 failing to undergo haemostatic testing despite having scores \nsuggestive of a significant bleeding tendency, and a further 14 due to an inability to confirm their diagnosis \nof endometriosis. Of the resultant 82 participants, the vast majority reported that their endometriosis had \nbeen diagnosed following investigation for pain (dysmenorrhoea/ generalised pelvic pain /dyspareunia \n/mittleschmerz). Three women reported infertility as a cause for investigation and 1 woman infertility and \ndysmenorrhoea. General population descriptors are shown in Table 1. \n \nTable 1 Population characteristics \n \n \n*During interview participants were asked: ‘Has a doctor ever told \nyou that you have heavy periods?’ and ‘Have you ever been \ntreated for heavy periods?’ Response was not always consistent – \nsome women reported having been treated for HMB but not told \nthat they had HMB. HMB was considered present if women had \nbeen treated for heaviness of periods. \n \nForty-one women underwent haemostatic testing following satisfaction of described criteria.  \nOne woman had a prolonged APTT, and all PT tests were normal. At least one vWF/FVIII test was found \nto be abnormal in 7/82 women giving a prevalence of  8.54% for abnormal haemostatic tests (95%CI \n3.50%-16.80%). Excluding 3 women with prolonged PFA-100 ® results in the context of having taken \nNSAIDs \n≤  14 days prior to testing, and 2 women who did not have PFA-100® tests performed, 10/77 \nwomen had at least one abnormal PFA-100® result giving a prevalence of 12.99% for abnormal PFA-100 \n®  results (95%CI 6.41-22.59%). Prolonged PFA-100® results were observed in several women in the \ncontext of NSAID  use ≤  14 days prior to testing, low haemotocrit and/or low platelet count, although in \nother women with the same factors, the results were normal. \n \nTherefore 14/79 women had at least one abnormal test, giving a prevalence of 17.72% abnormal \nhaemostatic tests (95% CI 10.04-27.94%). This, and all subsequent references to the overall prevalence for \nall tests combined excludes: 2 women for which PFA-100® was not performed and 1 woman whose only \nabnormal result was prolonged PFA-100® in the setting of having used NSAIDs  \n≤ 14 days prior to testing. \n \n All Participants \nN=82 \nAge (years) \nMean (Standard deviation) \nRange \n \n32.5 (7.45) \n18-48 \nGynaecologic History (%) \nHMB* \nFibroids \nPolyps \n \n41.46 \n24.39 \n12.20 \nMedian Age at Menarche (years) 13 \nPast History of Anaemia (%) 45.12 \nParity (%) \nNulliparous \nParous \n \n57/82 (69.51) \n25/82 (30.49) \nBleeding Score \nMedian \nRange (0-36) \n \n4 \n0-15 \nNo. of Symptoms \nMedian \nRange (0-12) \n \n2.5 \n0-6 \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\nBleeding score, number of symptoms, individual bleeding symptoms \nSpearman Rank Correlation provided strong evidence for a strongly positive relationship between bleeding \nscore and number of symptoms, and a weakly positive relationship between bleeding score and PBAC \nscores. Age appeared to be independent of each of bleeding score, number of symptoms, and PBAC scores \n(Table 2).  \n \nTable 2 Spearman Rank Correlation analysis \n \n P value Spearman Rank \nCorrelation coefficient \n \nAge and:   \nbleeding score 0.701  \nnumber of symptoms 0.914  \nPBAC ‘current’ score 0.682  \nPBAC ‘heaviest ever’ score 0.858  \n \nBleeding score and:   \nnumber of symptoms <0.0001 0.939 \nPBAC current score 0.002 0.374 \nPBAC ‘heaviest ever’ score 0.001 0.367 \n \nPBAC ‘current’ score and:   \nbleeding score 0.0021 0.3744 \nnumber of symptoms 0.0039 0.3537 \n \nPBAC ‘heaviest ever’ score \nand: \n  \nbleeding score 0.0007 0.3673 \nnumber of symptoms 0.0003 0.3946 \n \n \n \nThere did not appear to be evidence of a significant difference in age distribution between the women with \nnormal haemostasis and abnormal test/s(p=0.792), including when Mann-Whitney tests were performed \nseparately for vWF/FVIII (p=0.601) and PFA-100® (p=0.710). Parity had no effect on each of bleeding \nscore, number of symptoms, and PBAC scores when examined via the Mann-Whitney test. Likewise PBAC \nscore distribution of ‘current’ score versus heaviest ever’ score had no impact on normal versus abnormal \ntests for vWF tests, or all tests combined. \n \nWith respect to the clinical indicators of general bleeding tendency, the parous and nulliparous women \nwere analysed separately, since different cut-off were used. Utilising the ROC curves to evaluate the value \nof the bleeding score in nulliparous women (70% of the population), the bleeding score demonstrated \npromise as a tool to discriminate between those with normal and abnormal haemostasis, with very good \nArea Under Curve (AUC) point estimates (\n≥ 0.80) for all tests, vWF/FVIII and PFA-100® abnormalities. \nBleeding score had a poor performance amongst parous women. Number of symptoms had a fair \nperformance for all tests (AUC = 0.779 (95% CI 0.68-0.87)), vWF/FVIII tests (AUC 0.76 (95% CI 0.65-\n0.88)) and PFA-100® (AUC 0.77(95%CI0.67-0.87). \n \nFor vWF/FVIII abnormalities individual symptoms performed poorly, including several \nAUC point estimates around 0.5, which is considered a diagnostic test failure. For PFA-100®, HMB and \ncutaneous symptoms had good AUC point estimates, however 95% CI were quite wide, and the lower limit \nsurrounded 0.5 for 58 women with cutaneous symptoms. For all tests, HMB had a good point estimate for \nAUC, however 95% CI were quite wide. In the 25 parous participants PPH generated a good AUC point \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\nestimate for PFA-100® and all tests, however 95% CI were very wide and poor. \n \nUsing a ≥ 100 point threshold for ‘current’ and ‘heaviest ever’ PBAC scores respectively, \n53.85% and 76.54% of women had a score suggestive of HMB (Higham, O'Brien et al. 1990). Spearman \nRank Correlation provided strong evidence of a weak-moderate positive relationship between PBAC \n‘current’, ‘heaviest ever’ and bleeding score and number of symptoms. There was a significant difference \nin median PBAC score between ‘current’ and ‘heaviest ever’ PBAC, regardless of whether women were \ntaking medications which may have affected menstrual flow (combined oral contraceptive pill, \nlevonorgestrel intrauterine system, gonadotrophin releasing hormone analogues, tranexamic acid) at time of \n‘current’ PBAC or not.  PBAC score distribution did not appear to differ significantly between women with \nnormal and abnormal haemostatic result/s, except current PBAC for PFA-100® results (p=0.0415). As a \ndiagnostic test to discriminate between women with normal and abnormal haemostatic result/s both \n‘current’ and ‘heaviest ever’ PBAC had a poor performance, and only ‘current’ PBAC had a good point \nestimate for ROC AUC, although 95% CI were wide and poor (0.71, 0.56-0.87). \n \nDiscussion \nStudies have shown that women with HMB have a higher rate of mild bleeding disorders than the \nbackground population rate (Shankar, Lee et al. 2004). Although there is work that demonstrates that HMB \nis a risk factor for endometriosis (Vercellini, De Giorgi et al. 1997),  this study provides the first indication \nof the prevalence of bleeding test abnormalities in women who have a confirmed diagnosis of \nendometriosis. This finding is consistent with the epidemiological studies demonstrating an association \nbetween endometriosis and factors predisposing to greater exposure to menses (for example HMB) and the \npotential opportunity for increased retrograde menstruation (Cramer, Wilson et al 1986; Vercellini, De \nGiorgi et al.  1997; Missmer 2003), with subsequent implantation as endometriotic deposits (D'Hooghe and \nDebrock 2002) in accordance with Sampson’s hypothesis. \n \nThe overall prevalence of 17.72% of initial laboratory haemostatic abnormalities in the sample, is lower \nthan may be expected. Previously, the only other suggestion of an increased rate of bleeding disorders \namongst this population, has been the unverified self reported 30% rate of endometriosis reported in a \nsurvey of women with vWD (Kirtava, Drews et al. 2003). In contrast, particular effort was undertaken in \nthis study to have either histological proof or very clear supporting visual evidence of endometriosis (not \nsimply “pigmented spots”).  \n \nIn earlier studies, the prevalence of definitive bleeding disorders (compared to initial abnormal haemostatic \nscreening tests) was reduced following specialist evaluation, as occurred in Kadir’s study (Kadir, \nEconomides et al. 1998). In Kadir’s study, only 65% of those who initially had results suggestive of vWD \nwere defined as having the disorder on follow-up and repeat testing. Regardless, the rate of mild bleeding \ndisorders appears to be greater amongst women with endometriosis than the rate in the general population. \n \nThe detection of an underlying bleeding abnormality in a woman with endometriosis is pertinent to clinical \nmanagement on several levels. The use of NSAIDs to manage dysmenorrhoea/pelvic pain amongst women \nwith endometriosis is common. Although in healthy populations NSAIDs have been shown to ameliorate \nHMB compared to placebo (Lethaby, Augood et al 2007), in women with bleeding disorders they may be \ncounter-productive in addressing HMB (Lee, Chi et al 2006; Siegel and Kouides 2002).  \n \nFurther prospective study is required to determine if treatment strategies aimed at reducing menstrual blood \nloss would be beneficial in this group. Epidemiological data suggests that reduced menstrual loss associated \nwith increased parity (Eskenazi and Warner 1997) and the use of the oral contraceptive pill appear to \nreduce the risk for endometriosis (Vercellini, Eskenazi et al 2011).There is evidence from small studies to \nsuggest this may be effective in adolescents  where the rate of endometriosis in adolescents with \ndysmenorrhea was vastly lower than expected in a clinical setting where menstrual loss had been reduced \nwith the use of tranexamic acid or hormonal menstrual management prior to considering a laparoscopy to \ninvestigate their pain (Sachedina, Abu Bakar et al 2020). Long term follow-up of this cohort also revealed a \nlower than expected rate of endometriosis (Knox, Ong et al 2019). \n \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\nThere are some significant methodological limitations to this pilot study which warrant noting regarding \nthe detection of bleeding disorders.  \n \nFirstly, with respect to participants recruited, limitations relate to loss to follow-up of 9 women who \nsatisfied the criteria for ‘significant’ bleeding tendency however did not undergo haemostatic testing and \nwere subsequently excluded, which may have falsely limited the prevalence of haemostatic abnormalities. \nA further 14 were excluded due to an inability to confirm their endometriosis diagnosis (with their \noperative procedure performed at other hospitals and despite efforts, their notes could not be traced). \n \nSecondly, there were limitations with respect to the strength of the haemostasis testing approach. Funding \nlimitations restricted our capacity to undertake repeat haematological testing amongst both the women who \nunderwent initial haemostatic testing and returned abnormal results, as well as those with initial apparently \nnormal results. However, further and repeat testing were beyond the scope of this pilot study. It is \nacknowledged that a formal diagnosis (and implications of labelling) of a bleeding disorder for a patient \nrequires repeat testing and clinical evaluation by specialist haematological services. Women in the study \nwho returned abnormal initial haemostatic results, according to the criteria described, were referred on for \nfurther haematological evaluation. For women in this study who did not satisfy the described criteria for \nsuspicion of increased bleeding testing (following data acquired via the questionnaire and PBAC responses) \nhaemostatic testing was not performed. This was both pragmatic due to funding limitations as well as \nrecognizing that, for some, in the absence of significant bleeding history there may have been lack of \nenthusiasm to undergo testing and this may have adversely affected recruitment to this pilot study. Whilst \nmeasures to optimize the sensitivity of the screening tools for suspicious bleeding history were applied in \nthis study as described, it is possible that women who were not tested may nonetheless have also returned \nabnormal haemostatic tests had they been tested (hence the prevalence estimates for abnormal haemostatic \ntesting may be conservative). Future research involving testing of all women with endometriosis, whether \nthey have a suspicious bleeding history or not, as well as repeat and longitudinal testing, and with specialist \nhaematological input will be critical to confirm the results of this hypothesis-generating pilot study. \n \nRecognition of the abovementioned methodological limitations lead to the failure to submit this paper for \npublication at time of completion. More recently, realization by the senior author who has an ongoing \ninterest in this area that there have still been no studies to clarify the relationship between endometriosis \nand mild bleeding disorders, despite the significant passage of time since this study was undertaken, has \nlead to the decision to submit the paper now. This is in recognition that despite the data having a number of \nlimitations as described with candour, it represents the only effort to tackle the question regarding the \npotential relationship between mild bleeding disorders and endometriosis. Acknowledgement of this as a \npotential contributor to endometriosis may offer some women further options for treatment, the avoidance \nof potential surgical risks as well as the possibility of earlier intervention. As such, we would applaud \nfuture research with a larger participant sample, and which overcome the haematological testing limitations \nof this pilot study described, to further investigate the hypothesis-generating data reported here which may \nsuggest an increased bleeding tendency exists amongst some women with endometriosis.\n \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\n References \n \n1. Agarwal SK, Chapron C, Giudice LC, Laufer MR, Leyland N, Missmer SA, Singh SS, Taylor HS. \nClinical diagnosis of endometriosis: a call to action. Am J Obstet Gynecol. 2019; 220: 354-e1. \n2. Bendifallah S, Suisse S, Puchar A, Delbos L,Poilblanc M, Descamps P, Golfier F, Jornea L, Bouteiller \nD, Touboul C, Dabi Y, Darai E.  Salivary MicroRNA Signature for Diagnosis of Endometriosis. J Clin \nMed. 2022 11(3):612. \n3. Bowman M, Hopman WM, Rapson D, Lillicrap D, James P. The prevalence of symptomatic von \nWillebrand disease in primary care practice. J Thromb Haemost. 2010; 8: 213-216. \n \n4. Burney RO, Giudice LC. Pathogenesis and pathophysiology of endometriosis. Fertil Steril  2012;98:511–\n519. \n \n5. Chi C, Shiltagh N, Kingman CEC, Economides DL, Lee CA, Kadir RA. Identification and management \nof women with inherited bleeding disorders: a survey of obstetricians and gynaecologists in the United \nKingdom. Haemophilia. 2006;12:405-412. \n \n5. Cramer DW, Missmer SA.  The Epidemiology of Endometriosis (Endometriosis: emerging research and \nintervention strategies). Annals  New York Academy Sciences 2002; 955: 11-22. \n \n6. Cramer DW, Wilson E, Stillman RJ, Berger MJ, Beslisle S, Schiff I.  The relation of endometriosis to \nmenstrual characteristics, smoking and exercise. JAMA 1986; 255: 1904-1908. \n \n7. D'Hooghe TM, Debrock S. Endometriosis, retrograde menstruation and peritoneal inflammation in \nwomen and in baboons. Hum Reprod Update 2002; 8: 84-88. \n \n8. Dawson B, Trapp R . Research Questions about Two Separate or Independent Groups. Basic and \nClinical Biostatistics. 2004 B. Dawson and R. Trapp, McGraw-Hill Companies. \n \n9. Eskenazi B, Warner ML.  Epidemiology of Endometriosis. Obstet Gynecol Clinics 1997; 24(2): 235-257. \n \n10. Greaves M, Watson HG. Approach to the diagnosis and management of mild bleeding disorders. \nJournal of thrombosis and haemostasis : J Thromb Haemost 2007; 5 Suppl 1: 167-174. \n \n11. Halme J, Hammond MG, Hulka JF, Raj SG,Talbert LM. Retrograde menstruation in healthy women \nand in patients with endometriosis. Obstet Gynecol 1984; 64: 151-154. \n12. Hayward CPM, Harrison P, Cattaneo M, Ortel TL, Rao AK. Platelet function analyzer (PFA)-100 \nclosure time in the evaluation of platelet disorders and platelet function. Journal of Thrombosis and \nHaemostasis 2006; 4: 312-319. \n \n13. Higham JM, O'Brien PMS, Shaw RW. Assessment of menstrual blood loss using a pictorial chart. \nBritish Journal of Obstetrics and Gynaecology 1990; 97: 734-738. \n \n14. James, AH. More than menorrhagia: a review of the obstetric and gynaecological manifestations of \nbleeding disorders. Haemophilia 2005; 11: 295-307. \n \n15. James AH, Ragni MV, Picozzi VJ. Bleeding disorders in premenopausal women: (another) public \nhealth crisis for hematology? Hematology Am Soc Hematol Educ Program: 2006; 474-485. \n \n16. Jayasinghe Y,  Moore P, Donath S, Campbell J, Monagle P, Grover S.R. Bleeding disorders in \nteenagers presenting with menorrhagia. Australian New Zealand J Obstet Gynaecol 2005; 45: 439. \n \n17. Kadir RA, Economides DL, Sabin CA, Owens D, Lee CA. Frequency of inherited bleeding disorders in \nwomen with menorrhagia. The Lancet 1998; 351: 485-489. \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\n \n18. Kirtava A, Drews C, Lally C, Dilley A, Evatt B.  Medical, reproductive and psychosocial experiences \nof women diagnosed with von Willebrand's disease receiving care in haemophilia treatment centres: a case \ncontrol study. Haemophilia 2003; 9: 292-297. \n \n19. Koninckx PR, Ussia A, Adamyan L, Tahlak M, Keckstein J, Wattiez A, Martin DC. The epidemiology \nof endometriosis is poorly known as the pathophysiology and diagnosis are unclear Best Pract Res Clin \nObstet Gynaecol. 2021;71:14-26.  \n \n20. Kouides PA,  Phatak PD, Burkart, P Braggins C, Cox C, Bernstein Z, Belling L. Holmberg P, \nMaclaughlin W, Howard F. Gynaecological and obstetrical morbidity in women with type I von Willebrand \ndisease: results of a patient survey. Haemophilia 2000; 6(6): 643-648. \n \n21. Lee CA, Chi C, Pavord SR, Bolton-Maggs PHB, Pollard D, Hinchcliffe-Wood A, Kadir RA. The \nobstetric and gynaecological management of women with inherited bleeding disorders: A review with \nguidelines produced by a taskforce of UK Haemophilia Centre Doctors Organization. Haemophilia 2006; \n12: 301-336. \n \n22. Knox B, Ong Y, Bakar M, Grover SR. A longitudinal study of adolescents dysmenorrhoea into \nadulthood. Eur J Pediatr 2019; 178: 1325-1332. \n \n23. Lee CA, Kadir RA. von Willebrand Disease and Women's Health. Semin Hematol 2005; 42: 42-48. \n \n24. Lethaby A, Augood C, Duckitt K, Farquhar C. Nonsteroidal anti-inflammatory drugs for heavy \nmenstrual bleeding. Cochrane Database Syst Rev 2007; 4: CD000400. \n \n25. Lusher JM.  Systemic Causes of Excessive Uterine Bleeding. Semin Hematol 1999; 36(3 (S4): 10-20. \n \n26. Mannucci PM, Tuddenham EGD. The Hemophilias - From Royal Genes to Gene Therapy. N Engl J \nMed 2001; 344: 1773-1779. \n \n27. McPherson, R. and M. Pincus, Eds.  Henry's Clinical Diagnosis and Management by Laboratory \nMethods. Diagnostic accuracy. Receiver Operator Characteristic(ROC) Curves. 2007 Philadelphia, PA, \nSaunders Elsevier. \n \n28. Missmer S, Cramer D. The epidemiology of endometriosis. Obstet  Gynecol Clinics North Am 2003; \n30: 1-19. \n \n29. Mitri F, Casper RF. Endometriosis and mild bleeding disorders. Fertil Steril. 2015;103:886-887.e1.  \n \n30. Peyvandi F, Duga S, Akhavan S, Mannucci PM. Rare coagulation deficiencies. Haemophilia 2002; 8: \n308. \n \n31. Philipp CS, Faiz A, Dowling NF, Beckman M, Owens S, Ayers C, Bachmann G. Development of a \nscreening tool for identifying women with menorrhagia for hemostatic evaluation. Am J Obstet Gynecol \n2008; 198: 163.e1-163.e8. \n \n32. Rodeghiero F, Castaman G, Dini E. Epidemiological investigation of the prevalence of von \nWillebrand's disease. Blood 1987; 69: 454-459. \n \n33. Rodeghiero F, Castaman G, Tosetto A, Battle J, Baudo F, Cappelletti A, Casana P, De Bosch N, \nEikenboom JCJ, Federici AB, Lethagen S, Linari S, Srivastava A. The discriminant power of bleeding \nhistory for the diagnosis of type 1 von Willebrand disease: an international, multicenter study. Journal of \nThrombosis & Haemostasis, Blackwell Publishing Limited. 2005; 3: 2619-2626. \n \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint \n\n34. Rodeghiero F, Kadir RA, Tosetto A, James PD. Relevance of quantitative assessment of bleeding in \nhaemorrhagic disorders. Haemophilia 2008; 14(s3): 68-75. \n \n35. Rodeghiero F, Tosetto A, et al. How to estimate bleeding risk in mild bleeding disorders. Thromb \nHaemost 2007; 5(Suppl. 1): 157-166. \n \n36. Sachedina A, Abu Bakar M, Dunford AM, Morris A, Nur Azurah AG, Grover SR. Dysmenorrhea in \nyoung people: Experiences from a tertiary center with a focus on conservative management. J Obstet \nGynaecol Res. 2020;47: 352-358 \n \n37. Sadler JE. Von Willebrand disease type 1: a diagnosis in search of a disease. Blood 2003; 101: 2089-\n2093. \n \n38. Shankar M, Lee CA, Sabin CA, Economides DL, Kadir RA. von Willebrand disease in women with \nmenorrhagia: a systematic review. BJOG: An International J Obstet Gynaecol 2004; 111: 734-740. \n \n39. Siegel JE, Kouides PA. Menorrhagia from a haematologist's point of view. Part II: management. \nHaemophilia 2002; 8: 339-47 \n \n40. Sramek A, Eikenboom JC, Briet E, Vandenbroucke JP, Rosendaal FR. Usefulness of patient interview \nin bleeding disorders. Arch Intern Med 1995; 155: 1409-15. \n \n41. Surrey E, Carter CM, Soliman AM, Khan S, DiBenedetti DB, Snabes MC. Patient-completed or \nsymptom-based screening tools for endometriosis: a scoping review. Arch Gynecol Obstet 2017; 296: 153-\n165. \n \n42. Tosetto, A., F. Rodeghiero, Castaman G, Goodeve A, Federici AB, Batlle J, Meyer D, Fressinaud E et \nal. A quantitative analysis of bleeding symptoms in type 1 von Willebrand disease: results from a \nmulticenter European study (MCMDM-1 VWD). J Thromb Haemost 2006; 4: 766-773. \n \n43. Vercellini P, De Giorgi O, Aimi G, Panazza S, Uglietti A, Crosignani, PG. Menstrual Characteristics in \nWomen With and Without Endometriosis. Obstet Gynecol 1997; 90: 264-268. \n \n44. Vercellini P, Eskenazi B, Consonni D, Somigliana E, Parazzini F, Abbiati A,  Fedele L. Oral \ncontraceptives and risk of endometriosis: a systematic review and meta-analysis. Hum Reprod Update \n2011; 17: 159-170. \n \n45. Werner EJ, Broxson EH, et al. Prevalence of von Willebrand disease in children: A multiethnic study. J \nPediat 1993; 123: 893-898. \n \n46. Wykes CB, Clark TJ, et al. Accuracy of laparoscopy in the diagnosis of endometriosis: a systematic \nquantitative review. BJOG 2004; 111: 1204-12. \n \n \n . CC-BY-NC 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 preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.03.23286728doi: medRxiv preprint","source_license":"CC0","license_restricted":false}