UvA-DARE (Digital Academic Repository) Vasa previa and ... · Chapter 3 Incidence of and risk...

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Vasa previa and placenta associated complicationsDiagnosis and risk assessmentRuiter, L.

Publication date2017Document VersionFinal published versionLicenseOther

Link to publication

Citation for published version (APA):Ruiter, L. (2017). Vasa previa and placenta associated complications: Diagnosis and riskassessment.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an opencontent license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, pleaselet the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the materialinaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letterto: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. Youwill be contacted as soon as possible.

Download date:26 Mar 2021

https://dare.uva.nl/personal/pure/en/publications/vasa-previa-and-placenta-associated-complications(f142369f-3812-4947-a3ab-567e0764f569).html

UITNODIGING

voor het bijwonen van de openbare verdediging van het

proefschri�

Vasa previa and placenta associated complications

Diagnosis and risk assessment

op woensdag 22 november 2017 om 12.00 uur

in de Agnietenkapel van de Universiteit van Amsterdam Oudezijds Voorburgwal 231

te Amsterdam.

Na afl oop bent u van harte welkom bij de recep� e

ter plaatse.

Laura RuiterDodaarsweg 33

3897 LP Zeewolde06-49320999

[email protected]

Paranimfen:Chantal Diedrich

06-48506833

Marinka Bierma-Enthoven06-15662653

Vasa previa and placenta associated complications Diagnosis and risk assessment

Laura Ruiter

Vasa previa and placenta associated com

plications - diagnosis and risk assessment L

aura Ruiter

Ruiter_cover.indd Alle pagina's 6-10-2017 11:59:31



Laura Ruiter

Vasa previa and placenta associated complications – Diagnosis and risk assessment

Thesis, Academic Medical Centre, University of Amsterdam, The Netherlands

© Laura Ruiter, 2017

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means,

without written permission of the author.

Cover Remco Wetzels - www.remcowetzels.nl

Layout Nicole Nijhuis - Gildeprint

Yvonne Meeuwsen - www.yvonnemeeuwsen.nl

Printing Gildeprint, Enschede

ISBN 978-94-6233-759-6

Financial support of this thesis was kindly provided by BMA-Mosos, Chipsoft, Memidis Pharma BV and The University

of Amsterdam (UvA).



ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

prof. dr. ir. K.I.J. Maexten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapelop woensdag 22 november 2017, te 12:00 uur

door Laura Ruiter

geboren te Naarden

Promotiecommissie

Promotoren: Prof. Dr. E. Pajkrt AMC – UVAProf. Dr. B.W.J. Mol AMC – UVA

Copromotores: Dr. I.M. de Graaf AMC – UVA

Overige leden:Prof. Dr. M.W. Hollmann AMC – UVADr. A Kwee Universitair Medisch Centrum UtrechtDr. E.E.M. van Leeuwen AMC – UVAProf. Dr. D. Oepkes Universiteit van LeidenProf. Dr. J.A.M. van der Post AMC – UVA

Faculteit der Geneeskunde

CONTENTS

Chapter 1 General introduction, objectives and outline of this thesis 7

PART I VASA PREVIA 15

Chapter 2 Systematic review of accuracy of ultrasound in the diagnosis of vasa previa. 17 Ultrasound in Obstetrics and Gynecology. 2015 May;45(5):516-22

Chapter 3 Incidence of and risk indicators for vasa praevia: a systematic review. 35 British Journal of Obstetrics and Gynaecology. 2016 Jul;123(8):1278-87

Chapter 4 Screening, diagnostiek en beleid bij vasa previa – concept richtlijn 55

PART 2 PLACENTA ASSOCIATED COMPLICATIONS 65

Chapter 5 Predictors for emergency cesarean delivery in women with placenta previa. 67 American Journal of Perinatology. 2016 Dec;33(14):1407-1414

Chapter 6 Incidence and recurrence rate of placental abruption: 83 a longitudinal linked national cohort study in the Netherlands. American Journal of Obstetrics and Gynecology. 2015 Oct;213(4):573.e1-8

Chapter 7 Incidence and recurrence rate of hemorrhage postpartum and manual 99 removal of the placenta: a longitudinal linked national cohort study in the Netherlands. Submitted

Chapter 8 General discussion and implications for further research 115

Chapter 9 Nederlandse samenvatting 129 Portfolio 133

Publicatielijst 135 Dankwoord 137 Curriculum Vitae 141

CHAPTER 1

General introduction

Chapter 1

8

GENERAL INTRODUCTION

For women, to carry out a full term pregnancy and to deliver a healthy baby depends on several factors. The development and growth of the fetus is crucial and obviously cannot be accomplished without the placenta and umbilical cord. The placenta is a highly specialized organ of the pregnancy that supports the normal growth and development of the fetus. Development of the placenta is accurately regulated to ensure exchange of nutrients and waste products between the circulatory systems of the mother and fetus.1 The umbilical cord connects the placenta and the fetus and makes transport of these nutrients and waste products between them possible. Impaired functioning of the placenta is seen in many diseases such as pregnancy induced hypertension and pre-eclampsia. These complications can result in suboptimal growth of the fetus, preterm delivery and higher rates of neonatal intensive care unit admissions.2 But also morphological anomalies of the placenta and the umbilical cord can lead to maternal morbidity and poor perinatal outcome.

This thesis on anomalies of the umbilical cord and placenta will focus on vasa previa, placenta previa, placental abruption, postpartum hemorrhage and retained placenta.

A normally developed placenta consists of the basal plate (the maternal side of the placenta) and the chorionic plate (the fetal side of the placenta) and weighs about 600 grams at delivery in the term period of pregnancy.3, 4 The umbilical cord rises from the chorionic plate and in the majority of the pregnancies the origin is in the center part of the placenta. An abnormal umbilical cord insertion implies an insertion which is towards the placental edge and can be stratified as eccentric (>2cm from the placental edge), marginal (<2cm from the placental edge) or velamentous.

Vasa previaIn velamentous cord insertion the umbilical cord inserts outside the chorionic plate directly into the fetal membranes, it is seen in approximately 0.5% of all pregnancies.5, 6 In velamentous cord insertion, fetal vessels deriving from the umbilical cord run through the fetal membranes before entering the placenta and are partially unsupported by either the Wharton’s jelly -a specialized tissue that acts as supportive and protective structure substituting for the adventitia of the umbilical vessels- or placental tissue.6, 7 Pregnancies with a velamentous cord insertion are associated with adverse perinatal outcomes such as intra-uterine fetal demise, small for gestational age neonates and preterm delivery <37 weeks of gestation.5


9

CHAPTER

1Velamentous cord insertion can result in vasa previa when the fetal vessels that run through the membranes traverse the cervical internal os. Vasa previa can be divided into two subtypes; type I is a single lobed placenta with a velamentous cord insertion covering or close to the cervix and type II is a multi-lobed placenta with connecting fetal vessels running over the cervical internal os.8 (Figure 1a and b) Compression of these vessels by the presenting part may lead to fetal heart decelerations and bradycardia on the cardiotocogram.9, 10 Theoretically in case of rupture of membranes, vasa previa can rupture simultaneously, thus potentially causing fetal blood loss with serious neonatal morbidity or neonatal death.

Figure 1a. Type I vasa previa Figure 1b. Type II vasa previa

Placenta previaWhereas the majority of the placentas are located high on the anterior or posterior wall of the uterus, placental tissue can also cover the cervix, a condition which is called placenta previa. Placenta previa is seen in 0.4-0.8% of all pregnancies and is classified according to ultrasound imaging as complete (with placental tissue completely covering the internal cervical os) or incomplete (traditionally also called partial or marginal, with placental tissue reaching to but not covering the internal cervical os).11-13 Especially a complete placenta previa is associated with emergency delivery by Cesarean section and maternal morbidity such as antepartum hemorrhage and blood transfusion, need for hysterectomy postpartum and adverse perinatal outcomes as preterm delivery, anaemia and respiratory distress syndrome.11

Chapter 1

10

Placental abruptionPlacental abruption refers to the premature dehiscence of a normally inserted placenta from the uterine wall. It is associated with significant perinatal morbidity and accounts for 12% of all perinatal deaths in the United States.14 Diagnosis of placental abruption is difficult, since the clinical presentation includes not very specific symptoms such as vaginal bleeding and abdominal pain. Early diagnosis of placental abruption can potentially be obtained by intensive surveillance and may improve the outcome of children involved in these pregnancies, by closely monitoring women who are at risk.15 Advanced maternal age, multiparity, chronic hypertension, smoking, cocaine use, previous Cesarean delivery, uterine surgery and short interpregnancy interval are reported risk factors for placental abruption.16-20

Retained placenta and postpartum hemorrhageIn addition to prenatal complications regarding the placenta, the placenta can also cause postpartum complications. In the normal situation the placenta detaches from the uterine wall spontaneously and is delivered shortly after delivery of the neonate. Although there is no universal consensus for the length of time allotted, several guidelines suggest to use 30 minutes after delivery of the neonate as the length of time before an intervention is advised.21 Interventions can involve manual or operative removal of the placenta and can potentially increase postpartum hemorrhage and infections. This is of importance, since postpartum hemorrhage is still one of the leading causes of maternal morbidity and mortality in developed and third world countries despite improvements in prevention and management.22, 23

BackgroundAt the time the research in this thesis was initiated, the literature on vasa previa was mainly restricted to case-reports, reviews and small case-series. The focus in these studies particularly lay on the outcome of vasa previa, which appeared to depend on prenatal diagnosis.8, 24, 25 Unfortunately since vasa previa is a rather rare complication, publication bias seemed to be a significant problem with mostly publication of case reports reporting adverse neonatal outcome such as fetal demise in ruptured vasa previa. Studies evaluating the prenatal diagnosis of vasa previa and its outcome objectively were not readily available. It has been hypothesized that prenatal recognition of vasa previa by ultrasound screening enables elective delivery of the fetus by Cesarean section, thus avoiding potential neonatal morbidity or fetal demise.25, 26 Different researchers and patient associations advocated to start a screening program on vasa previa for all pregnant women based on the available literature.

Placenta associated complications such as placenta previa, placental abruption and postpartum hemorrhage due to retained placenta have been studied previously to a greater


11

CHAPTER

1or lesser extent. But the consequences for women regarding future pregnancies after suffering placental abruption or severe postpartum hemorrhage due to retained placenta were not clear. Consequently, prenatal and perinatal care in women with a history of placental abruption, severe postpartum hemorrhage and/or retained placenta is not clearly defined in guidelines and therefore varies worldwide.15 In addition, for women with placenta previa several recommendations can be found regarding timing of delivery (most often a preterm elective Cesarean section) and administration of antenatal corticosteroids for fetal lung maturation, but there still is a lack of consensus.27

Objectives and outline of this thesisThe objectives of this thesis were to systematically assess the established screening criteria for the development of a screening program regarding vasa previa. We also identified factors that predict the risk of emergency delivery in women with placenta previa and we investigate the consequences of a history of placental abruption, severe postpartum hemorrhage and retained placenta for potential future pregnancies.

Part 1. Vasa previaIn the first part of this thesis we focus on vasa previa and investigate the possibilities for a potential screening strategy.

Chapter 2 What is the diagnostic accuracy of ultrasound in the prenatal diagnosis of vasa previa?

Chapter 3 What is the incidence of vasa previa and what are risk indicators for vasa previa?Chapter 4 Screening, diagnostiek en beleid bij vasa previa – concept richtlijn. Part 2. Placenta associated complicationsIn the second part of this thesis we focus on placenta previa, placental abruption and retained placenta with hemorrhage postpartum.

Chapter 5 What are risk factors for emergency delivery in women with placenta previa?Chapter 6 What is the incidence and recurrence rate of placental abruption?Chapter 7 What is the incidence and recurrence rate of severe postpartum hemorrhage

and/or retained placental tissue?

Chapter 8 General discussion and implications for further research

Chapter 1

12

REFERENCES

1. Gude NM, Roberts CT, Kalionis B, King RG. Growth and function of the normal human placenta. Thrombosis research. 2004;114(5-6):397-407.

2. Habli M, Levine RJ, Qian C, Sibai B. Neonatal outcomes in pregnancies with preeclampsia or gestational hypertension and in normotensive pregnancies that delivered at 35, 36, or 37 weeks of gestation. American journal of obstetrics and gynecology. 2007 Oct;197(4):406.e1-7.

3. Janthanaphan M, Kor-Anantakul O, Geater A. Placental weight and its ratio to birth weight in normal pregnancy at Songkhlanagarind Hospital. J Med Assoc Thai. 2006 Feb;89(2):130-7.

4. Lurie S, Feinstein M, Mamet Y. Human fetal-placental weight ratio in normal singleton near-term pregnancies. Gynecol Obstet Invest. 1999;48(3):155-7.

5. Esakoff TF, Cheng YW, Snowden JM, Tran SH, Shaffer BL, Caughey AB. Velamentous cord insertion: is it associated with adverse perinatal outcomes? The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2015 Mar;28(4):409-12.

6. Eddleman KA, Lockwood CJ, Berkowitz GS, Lapinski RH, Berkowitz RL. Clinical significance and sonographic diagnosis of velamentous umbilical cord insertion. American Journal of Perinatology. 1992 3/1992;9(2):123-6.

7. Kulkarni ML, Matadh PS, Ashok C, Pradeep N, Avinash T, Kulkarni AM. Absence of Wharton’s jelly around the umbilical arteries. Indian J Pediatr. 2007 8/2007;74(8):787-9.

8. Catanzarite V, Maida C, Thomas W, Mendoza A, Stanco L, Piacquadio KM. Prenatal sonographic diagnosis of vasa previa: ultrasound findings and obstetric outcome in ten cases. Ultrasound in Obstetrics & Gynecology. 2001 8/2001;18(2):109-15.

9. Naftolin F, Mishell DR, Jr. Vasa previa. Report of 3 cases. Obstetrics & Gynecology. 1965 10/1965;26(4):561-5.

10. Antoine C, Young BK, Silverman F, Greco MA, Alvarez SP. Sinusoidal fetal heart rate pattern with vasa previa in twin pregnancy. Journal of Reproductive Medicine. 1982 5/1982;27(5):295-300.

11. Dola CP, Garite TJ, Dowling DD, Friend D, Ahdoot D, Asrat T. Placenta previa: does its type affect pregnancy outcome? Am J Perinatol. 2003 Oct;20(7):353-60.

12. Tuzovic L. Complete versus incomplete placenta previa and obstetric outcome. Int J Gynaecol Obstet. 2006 May;93(2):110-7.

13. Vergani P, Ornaghi S, Pozzi I, Beretta P, Russo FM, Follesa I, et al. Placenta previa: distance to internal os and mode of delivery. American journal of obstetrics and gynecology. 2009 Sep;201(3):266 e1-5.

14. Ananth CV, Wilcox AJ. Placental abruption and perinatal mortality in the United States. Am J Epidemiol. 2001 Feb 15;153(4):332-7.

15. Rasmussen S, Irgens LM, Albrechtsen S, Dalaker K. Women with a history of placental abruption: when in a subsequent pregnancy should special surveillance for a recurrent placental abruption be initiated? Acta obstetricia et gynecologica Scandinavica. 2001 Aug;80(8):708-12.

16. Ananth CV, Savitz DA, Luther ER. Maternal cigarette smoking as a risk factor for placental abruption, placenta previa, and uterine bleeding in pregnancy. Am J Epidemiol. 1996 Nov 1;144(9):881-9.

17. Ananth CV, Savitz DA, Williams MA. Placental abruption and its association with hypertension and prolonged rupture of membranes: a methodologic review and meta-analysis. Obstetrics and gynecology. 1996 Aug;88(2):309-18.

18. Getahun D, Oyelese Y, Salihu HM, Ananth CV. Previous cesarean delivery and risks of placenta previa and placental abruption. Obstetrics and gynecology. 2006 Apr;107(4):771-8.


13

CHAPTER

119. Hulse GK, Milne E, English DR, Holman CD. Assessing the relationship between maternal cocaine

use and abruptio placentae. Addiction. 1997 Nov;92(11):1547-51.20. Tikkanen M, Nuutila M, Hiilesmaa V, Paavonen J, Ylikorkala O. Prepregnancy risk factors for

placental abruption. Acta obstetricia et gynecologica Scandinavica. 2006;85(1):40-4.21. Coviello EM, Grantz KL, Huang CC, Kelly TE, Landy HJ. Risk factors for retained placenta. American

journal of obstetrics and gynecology. 2015 Dec;213(6):864 e1- e11.22. World Health Organization. WHO Recommendations for the prevention and treatment of

postpartum haemorrhage. . Geneva (Switzerland); 2012.23. Mathai M, Gulmezoglu AM, Hill S. Saving womens lives: evidence-based recommendations for

the prevention of postpartum haemorrhage. Bulletin of the World Health Organization. 2007 Apr;85(4):322-3.

24. Lee W, Lee VL, Kirk JS, Sloan CT, Smith RS, Comstock CH. Vasa previa: prenatal diagnosis, natural evolution, and clinical outcome. Obstetrics & Gynecology. 2000 4/2000;95(4):572-6.

25. Smorgick N, Tovbin Y, Ushakov F, Vaknin Z, Barzilay B, Herman A, et al. Is neonatal risk from vasa previa preventable? The 20-year experience from a single medical center. Journal of Clinical Ultrasound. 2010 3/2010;38(3):118-22.

26. Seince N, Carbillon L, Perrot N, Uzan M. Various Doppler sonographic appearances and challenges in prenatal diagnosis of vasa praevia. Journal of Clinical Ultrasound. 2002 9/2002;30(7):450-4.

27. Balayla J, Wo BL, Bedard MJ. A late-preterm, early-term stratified analysis of neonatal outcomes by gestational age in placenta previa: defining the optimal timing for delivery. The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2015;28(15):1756-61.

PART IVASA PREVIA

CHAPTER 2

Systematic review of accuracy of ultrasound in the

diagnosis of vasa previa

Ruiter L Kok N

Limpens J Derks JB

de Graaf IMMol BWJ

Pajkrt E

Ultrasound in Obstetrics and Gynecology. 2015 May;45(5):516-2

Chapter 2

18

ABSTRACT

Objective Vasa previa is an obstetric complication in which the fetal blood vessels lie outside the chorionic plate in close proximity to the internal cervical os. In women with vasa previa, the risk of rupture of these vessels is increased, thus potentially causing fetal death or serious morbidity. Our objective was to assess the accuracy of ultrasound in the prenatal diagnosis of vasa previa.

Methods We searched MEDLINE, EMBASE, the Cochrane Library and PubMed for studies on vasa previa. Two reviewers independently selected studies on the accuracy of ultrasound in the diagnosis of vasa previa. Studies were scored on methodological quality using the Quality Assessment of Diagnostic Accuracy Studies tool (QUADAS-2). Data on sensitivity and specificity were subsequently extracted.

Results The literature search revealed 583 articles of which two prospective and sixretrospective cohort studies were eligible for inclusion in the qualitative analysis. All studies documented methods suitable for the prenatal diagnosis of vasa previa. Four out of the eight studies used transvaginal ultrasound (TVS) for primary evaluation, while the remaining four studies used transabdominal ultrasound and performed a subsequent TVS when vasa previa was suspected. The QUADAS-2 tool reflected poor methodology in six of the eight included studies, and prenatal detection rates varied from 53% (10/19) to 100% (total of 442,633 patients, including 138 cases of vasa previa). In the two prospective studies (N=33,795, including 11 cases of vasa previa) transvaginal color Doppler performed during the second trimester detected all cases (sensitivity 100%) for a specificity of 99.0-99.8%.

Conclusion The accuracy of ultrasound in the diagnosis of vasa previa is high, when performed transvaginally in combination with color Doppler.

Systematic review of accuracy of ultrasound in the diagnosis of vasa previa

19

CHAPTER

2

INTRODUCTION

Vasa previa is a complication of pregnancy in which fetal blood vessels lie outside the chorionic plate, in close proximity to the internal cervical os. The estimated incidence of VP has been put at between 1:1200 and 1:5000, but recent research that it might not be as rare as once thought. 1-3 Vasa previa can be divided into two subtypes; Type I is the occurrence of a single lobed placenta with a velamentous cord insertion and Type II is a multi-lobed placenta with connecting vessels running over the internal cervical os. 4 Normally, the umbilical cord is protected by placental tissue or Wharton’s jelly, a specialized tissue that acts as a supportive and protective structure substituting for the adventitia of the umbilical vessels. 5 In vasa previa this protection is absent and compression of the vessels by the presenting part may lead to fetal heart decelerations and bradycardia on the cardiotocography. 6-8 In cases of rupture of membranes, vasa previa fetal vessels can rupture simultaneously, thus potentially causing fetal blood loss and serious neonatal morbidity and mortality. 9 It has been hypothesized that prenatal recognition of vasa previa by ultrasound screening enables elective delivery of the fetus by Cesarean section, thus avoiding potential fetal demise or neonatal morbidity. 10 Nowadays, the majority of pregnant women undergo several scans during pregnancy. Locating and documenting the umbilical cord insertion site can be done, but is not routinely carried out. It has been proposed recently that a standard evaluation of the umbilical cord insertion site be included in the second-trimester ultrasound examination to detect vasa previa prenatally. 9, 11, 12 The diagnostic performance of ultrasound in the prenatal detection of VP, however, is unknown, and this information is needed before one can decide whether or not to include screening for vasa previa in standardized prenatal care. The aim of our study was, therefore, to investigate the accuracy of ultrasound in the diagnosis of vasa previa.

METHODS

A medical librarian (J.L.) performed a comprehensive search of MEDLINE (OVID, from 1948), EMBASE (OVID, from 1947). The Cochrane Library, including the Cochrane Central Register of Controlled Trials (CENTRAL, from inception), PubMed (the publisher[sb]-fraction which contains publications ahead of print, not yet included in OVID MEDLINE) and ongoing Trial registers (http://clinicaltrials.gov/) until February 2014. Language restrictions were not applied and animal studies were excluded. The search strategy consisted of Subject Headings (MeSH, SH) and words in title and abstract for vasa previa or its synonyms (vasa previa Type I). In addition we searched broadly for diagnostic imaging and screening of abnormal umbilical cord location (vasa previa Types II and III); see Appendix S1 for entire MEDLINE search).

Chapter 1

20

The search included an iterative process for each database, to refine the search strategy through incorporation of new search terms as new relevant citations were identified, i.e. by checking reference lists and citing articles using ISI Web of Science. The bibliographic records retrieved were downloaded and imported into Reference Manager ® software version 12.0 (Carlsbad, CA, USA) to exclude duplicates, and the search results were stored and analyzed. Studies were selected in a two-stage process. First, two reviewers (L.R., N.K.) scrutinized titles and abstracts of all studies. Studies were considered eligible if they described ultrasound and its accuracy in the prenatal diagnosis of vasa previa, which was defined as fetal vessels in close proximity of the internal cervical os. Decisions on final inclusions and exclusions were made after the full-text articles on first selection had been read independently (L.R., N.K.) and examined in more detail. Disagreements about eligibility were resolved by discussion. We included prospective and retrospective cohort studies and case-control studies. The index test consisted of an ultrasound evaluation performed during pregnancy with the specific aim of detecting vasa previa. All studies included in the analysis performed at least one second-trimester scan. The reference standard for confirmation of vasa previa after delivery was macroscopic observation of the placenta and umbilical cord insertion site by the caregiver. Two reviewers (L.R., N.K.) scored all included studies independently on methodological quality with the use of the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2).13 This tool is recommended for assessment of the risk of bias and the applicability of primary diagnostic accuracy studies in systematic reviews. QUADAS-2 consists of four key domains; ‘patient selection’, ‘index test’, ‘reference standard’ and ‘flow and timing’. All four domains are rated in terms of risk of bias, and the last three are also rated in terms of applicability to the review question. 13 The quality items assessed were study design and the conduct and analysis of all included studies. Elements considered to be associated with a risk of bias were the use of retrospective data, non-consecutive patient enrollment, lack of blinding for the outcome of the reference standard at the time of the index test and a discrepancy in the number of patients who received the index test and the reference standard. Subsequently, both reviewers (L.R., N.K.) extracted clinical study characteristics independently using a predesigned data extraction form, including: country of investigation, period of data collection, number of cases of vasa previa, study population, method of ultrasound, gestational age at diagnosis, index test and reference standard. In addition, data on prenatal detection rates with sensitivity and specificity, were extracted.

Statistical analysisWe constructed two-by-two tables, cross-classifying the outcome of the index test against the outcome of the reference standard. Sensitivity and specificity were calculated, as well as positive and negative predictive value. Authors were contacted for additional data if it was


21

CHAPTER

2

not possible to create two-by-two tables. A random effects model was used for statistical pooling of the data and to present pooled date with 95% CIs. Further, an I2 test was applied to assess heterogeneity among the included studies and to calculate the area under the summary ROC curve (AUC) to measure the accuracy of ultrasound.

RESULTS

The initial search resulted in 583 unique articles, of which 563 articles were excluded after reading the title and abstract (Figure 1). The full text of the remaining 20 articles was examined in more detail and 12 articles were excluded; five did not document prenatal detection of vasa previa, but recorded outcome 2, 14, 15 , management 15 , cost-effectiveness of screening 16 and the clinical significance of velamentous cord insertion 17, one article was excluded as it reported on placental localisation in the first trimester, two articles were review. 12, 18, 19 and three articles were conference abstracts of studies not yet published. 20-22 One article reported on additional cases of a cohort previously described by Lee et al. but both articles were included in the analysis after exclusion of the double cases. 23, 24 Authors were contacted but no additional data were obtained. Ultimately, we included eight articles. Table 1 provides an overview of the included studies and their characteristics. A total of 442 663 women were included of which there were 138 cases of vasa previa. The number of vasa previa cases per study varied from 125 to 60 23 and all of the eight included studies were cohort studies. Results of quality assessment of the included studies are summarized in Figure 2. Data collection was prospective in two (25%) studies and sampling of patients was consecutive in five (63%) studies. Patients included in all eight studies were likely to match our target population. In all studies, the outcome of the reference standard was unknown at the time of the index test 26 and all studies documented the methods applied for prenatal diagnosis of vasa previa. Five studies documented the umbilical cord insertion site in all pregnancies and one study merely recorded this in twin pregnancies. 23 Two studies systematically reported placental location in all investigated pregnancies, while five studies systematically visualised and documented the lower uterine region or internal cervical os. 4, 11, 23, 24, 27

Transvaginal ultrasound (TVS) was the primary method used to diagnose vasa previa in four studies and was performed after a transabdominal scan (TAS) in three studies. One study investigated umbilical cord insertion site by TAS in the-mid trimester and patient with an inconclusive diagnosis were scheduled for TVS evaluation between 30 and 36 weeks’ gestation.25 All studies reported the use of color or pulsed Doppler for the diagnosis of vasa previa.

Chapter 2

22

In five of the eight studies11, 24, 25, 27, 28, gestational age at the time of ultrasound was reported, ranging from from 18+0 to 26+6 weeks. Three studies documented early third-trimester ultrasound scans to ascertain the diagnosis of (suspected) vasa previa. 4, 11, 25

Statistical pooling of the data on predictive parameters for diagnostic accuracy could not be performed, because of insufficient data owing to heterogeneity. In two prospective cohort studies4, 25 detection of velamentous cord insertion and vasa previa were the primary objectives. Nomiyama et al.25 reported that the umbilical cord insertion site was successfully visualized in 99.8% (586/587) of the fetuses and identification of a velamentous cord insertion site by ultrasound evaluation had a sensitivity of 100% (5/5),

Figure 1. PRISMA flowchart


23

CHAPTER

2

Tabl

e 1.

Cha

ract

eris

tics o

f the

incl

uded

stud

ies o

n de

tecti

on o

f vas

a pr

evia

(VP)

by

tran

svag

inal

(TVS

) and

tran

sabd

omin

al (T

AS) u

ltras

ound

Stud

yTy

pe a

nd se

tting

Incl

usio

n cr

iteria

Popu

latio

n (n

)Ca

ses o

f VP

(n)

Mat

erna

l ag

e (y

ears

)In

dex

test

Refe

renc

e st

anda

rd

Baul

ies

et

al.(3

3)20

07

Retr

ospe

ctive

co

hort

, sin

gle

cent

er, S

pain

All d

eliv

erie

s be

twee

n 20

00

and

2005

12 0

639

34 (2

8-40

)TV

S on

em

pty

blad

der a

t 20-

22 w

eeks

; col

or D

oppl

er fo

r co

nfirm

ation

of d

iagn

osis

Path

olog

ical

exa

min

ation

of

plac

enta

and

mem

bran

es

in fo

ur c

ases

; pla

cent

al

mor

phol

ogic

exa

min

ation

afte

r de

liver

y

Bron

stee

n et

al

.(44)

201

3

Retr

ospe

ctive

co

hort

, sin

gle

cent

er, U

SA

All d

eliv

erie

s of

con

firm

ed V

P be

twee

n 19

90-2

010

Excl

uded

: pati

ents

with

VP

with

pre

nata

l car

e an

d de

liver

y in

ano

ther

cen

ter

182

554

60U

nkno

wn

TAS

on fu

ll bl

adde

r; TV

S w

ith

colo

r Dop

pler

whe

n ne

eded

; ge

stati

onal

age

at d

iagn

osis

unkn

own

Deliv

ery

info

rmati

on, m

edic

al

reco

rd o

r disc

harg

e re

cord

Cata

nzar

ite e

t al

.(8)

2001

Pros

pecti

ve

coho

rt, s

ingl

e ce

nter

, USA

Patie

nts

diag

nose

d w

ith V

P by

ul

tras

ound

33 2

0810

31

TAS

with

col

or D

oppl

er o

n fu

ll bl

adde

r, be

fore

26w

eeks

of

gest

ation

; TVS

whe

n ce

rvix

co

uld

not b

e se

en b

y TA

S; fr

om

1998

: rou

tine

colo

r sw

eep

over

lo

wer

ute

rine

segm

ent

Clin

ical

or p

atho

logi

cal

exam

inati

on o

f pla

cent

a an

d m

embr

anes

Has

egaw

a et

al

.(48)

2010

Retr

ospe

ctive

co

hort

, sin

gle

cent

er, J

apan

All d

eliv

erie

s be

twee

n 20

06-

2009

4 53

210

34±4

.0Se

cond

-trim

este

r TVS

with

col

or

Dopp

ler

Not

repo

rted

Kand

a et

al.(

47)

2011

Retr

ospe

ctive

co

hort

, sin

gle

cent

er, J

apan

Patie

nts

with

VP

in m

edic

al

reco

rd b

etw

een

2002

-200

75

131

1031

.5±8

.4O

bser

vatio

n of

inte

rnal

cer

vica

l in

tern

al o

s by

TVS

with

col

or

Dopp

ler 2

0 an

d 25

wee

ks

Clin

ical

exa

min

ation

of p

lace

nta

and

mem

bran

es

Lee

et a

l.(24

)20

00

Retr

ospe

ctive

co

hort

, sin

gle

cent

er, U

SA

Patie

nts

with

sus

pect

ed V

P in

a c

ompu

teriz

ed d

atab

ase

betw

een

1991

-199

893

874

18U

nkno

wn

TAS

in e

arly

third

trim

este

r; ev

alua

tion

by T

VS w

ith c

olor

Do

pple

r whe

n su

spic

ion

of V

P

Clin

ical

or p

atho

logi

cal

exam

inati

on o

f pla

cent

a an

d m

embr

anes

Nom

iyam

a et

al

.(45)

1998

Pros

pecti

ve

coho

rt, s

ingl

e ce

nter

, Jap

an

Fetu

ses

scan

ned

at

18-2

0wee

ks’ b

etw

een

1993

-19

9658

71

Unk

now

n

TAS

with

col

or D

oppl

er in

se

cond

and

third

trim

este

rs;

TVS

if co

rd in

serti

on n

ot s

een

in

third

trim

este

r

Path

olog

ical

exa

min

ation

of

plac

enta

and

mem

bran

es

Smor

gick

et

al.(2

5)20

10

Retr

ospe

ctive

co

hort

, sin

gle

cent

er, I

srae

l

Patie

nts

with

VP

in m

edic

al

reco

rd fr

om 1

988-

2007

Excl

uded

: cas

es o

f ve

lam

ento

us c

ord

inse

rtion

w

ithou

t VP

110

684

1933

(22-

46)

TVS

with

col

or D

oppl

er

betw

een

15 a

nd 3

3 w

eeks

Clin

ical

exa

min

ation

of p

lace

nta

and

mem

bran

es

Mat

erna

l age

dat

a pr

ovid

ed a

s m

edia

n (r

ange

) or m

ean

±SD.

Chapter 2

24

a specificity of 99.8% (580/581), a positive predictive value of 83% (5/6) and a negative predictive value of 100% (580/580). 25

Catanzarite et al.4 diagnosed 11 cases (0.03%) of vasa previa by ultrasound among 33 208 women. Ten cases were confirmed at delivery while the eleventh case was a false-positive diagnosis that appeared to be a placenta previa at delivery. No standard outcome of all scanned pregnancies was available thus only specificity (100%) could be calculated in this study.

Table 2. Prenatal detection rates of vasa previa by ultrasound

Study

Cases of vasa previa Prenatal diagnosis Diagnosed after delivery

Baulies et al. (33) 9 9 (100) -

Bronsteen et al. (44) 60 56 (93) 4 (7)

Catanzarite et al. (8) 10 10 (100) -

Hasegawa et al. (48) 10 10 (100) -

Kanda et al. (47) 10 9 (90) 1 (10)

Nomiyama et al. (45) 1 1 (100) -

Smorgick et al. (25) 19 10 (53) 9 (47)

Data given as n or n (%). The study of Lee et al.(24) was not included in this table as information regarding prenatal diagnosis of vasa previa was presented in the study of Bronsteen et al.(44)

Figure 2. Results of Quality Assessment for Diagnostic Accuracy Studies


25

CHAPTER

2

In the remaining retrospective cohort studies, all reported on the number of prenatally diagnosed cases of vasa previa. The prenatal detection rate varied from 53 to 100%, as shown in table 2, with four studies diagnosing all cases of vasa previa prenatally. 4, 11, 25, 29 In contrast, Smorgick et al.28 diagnosed only 53% (10/19) of cases with vasa previa prenatally. In the nine cases that remained undetected, the umbilical cord insertion site had not been evaluated and only TAS had been performed. Two studies were able to review images of missed cases; Catanzarite et al.4 re-evaluated the stored images of these cases, and structures suggestive for vasa previa were discovered; Bronsteen et al.23 reviewed retrospectively the digital clips and videos of the first ultrasound in which they missed 13/58 cases initially. Four of those cases had recordings of the lower uterine segment suggestive of vasa previa, six cases had no suspicious areas despite adequate visualization and three had no suspicious areas but also had inadequate visualization for an appropriate evaluation. Little can be said about the false-positive rates of ultrasound in the diagnosis of vasa previa, owing to a lack of information. Bronsteen et al.23 reported five false-positive cases, but did not elaborate on these further, Nomiyama et al.25 reported one false positive case and Catanzarite et al.4 reported one false-positive case within their cohort.

DISCUSSION

In this systematic review we found that the accuracy of ultrasound in the prenatal detection of vasa previa is high, when performed transvaginally and combined with color Doppler. The median prenatal detection rate was 93%, and the specificity was between 99 and 100% in the present studies; however, the quality of the available studies is relatively poor. Prenatal diagnosis of vasa previa was made most frequently at 18-26 weeks’ gestation. It is useful to evaluate the placental cord insertion by TAS in the second trimester. An increase in missed cases of vasa previa is seen when ultrasound evaluation does not involve color Doppler, is transabdominal and/or is performed only in the third trimester. The optimal timing for the detection of VP has been the subject of discussion and evidence on earlier detection than the second trimester is available. Hasegawa et al.30 visualized placental cord insertion site in 93.5% of their patients at the time of the nuchal translucency scan. In another study31 the group found an association between vasa previa and placental cord insertion in the lower third of the uterus in the first trimester. Ultrasound evaluation of the cervical region in the third trimester may be limited owing to the presence of the fetal presenting part or a scar from a previous Cesarean section. 27

Nomiyama et al.25 and Sepulveda et al.32 reported an average of 20-30 additional seconds needed to evaluate the placental cord insertion and, in 95% of the cases the insertion site

Chapter 2

26

was seen within 1 minute. The occurrence of linear or tubular structures near the internal cervical os should arouse suspicion and mandate referral to a prenatal diagnostic centre for further evaluation, given the suspected unfavourable natural course of vasa previa. 2

Awareness on the subject of vasa previa seems to increase the prenatal detection rate when we compare the prospective study of Nomiyama et al.25 included in our review with that of Eddleman et al.17, in which a cohort of cases with velamentous cord insertion was reviewed retrospectively and the authors reported that none of the cases was diagnosed prenatally and, moreover, three cases of vasa previa had been missed. This is further supported by the study of Smorgick et al.28 included in our review, in which, despite the small number of cases (n=19), an increase in prenatal detection rate of vasa previa over two consecutive decades, from 25% in 1988-1997 to 60% in 1998-2007, was documented. Several potential limitations of our study should be pointed out. Despite our broad literature search, we found only two prospective studies. This may reflect the fact that vasa previa is a rare complication that may result in a lack of attention to the subject. Otherwise, it might reflect the difficulty of performing prospective cohort studies evaluating relatively rare events. Because of heterogeneity between the studies, it was not possible to perform a meta-analysis on sensitivity and specificity of ultrasound as a diagnostic tool for vasa previa. Consequently, most knowledge comes from retrospective studies carrying a substantial risk of selection bias. Verification bias is another concern, whereby performance of the reference test depends on the result of the index test.33 Verification bias may introduce a sensitivity estimate that is too high. Moreover, pathological examination cannot provide precise information on the exact position of the fetal vessels with respect to the internal os at the time of the delivery. It is important to bear in mind that every false-positive case means a potential unnecessary Cesarean section, with all its consequences. Many studies reported only postnatally confirmed cases of vasa previa and therefore no sensitivity and specificity could be determined. 11, 23, 24 Although it seems that the majority of cases of vasa previa have an unfavourable outcome, some remain uncomplicated. Most probably, some undetected cases of vasa previa remain uncomplicated and will not be registered as a false negative. Since there is no place for the documented universal macroscopic evaluation of all placentae, this will remain a limitation of future studies. 34 Lack of documentation on evaluation of overall perinatal death is another limitation that could result in underestimation of perinatal mortality due to vasa previa and, at the same time, an overestimation of the predictive capability of TVS in the diagnosis of vasa previa. 11,

25, 29 In the Netherlands, Huidekoper35 found ruptured vasa previa as the cause of bleeding during labor in 16 (0.05%) pregnancies when systematically investigating 30.000 placentae in 1972. Unfortunately, the outcomes of the children involved in these pregnancies were not reported. To assess how many perinatal deaths are attributable to vasa previa, it would be necessary to investigate all cases of perinatal death due to exsanguination over a certain


27

CHAPTER

2

period. As mentioned above, there is no evidence that every case of vasa previa results in fetal demise when it remains undiagnosed. A final issue to consider is the effectiveness of Cesarean section to prevent hemorrhage of the fetal vessels once vasa previa has been diagnosed. It is difficult to see how randomized clinical trials on the subject could be carried out, therefore it is important to collect data on the outcome of vasa previa in case of vaginal delivery. Also, we should realize that other studies involving interventions in vaginal delivery such as artificial rupture of membranes or induction of labor are usually underpowered to address complications such as bleeding from vasa previa.36, 37 Indeed, historical cohort studies conducted before the era of ultrasound may help us to assess the true importance of undiagnosed vasa previa in perinatal outcomes. In conclusion, as a first step in the evaluation of potential screening for vasa previa, we conclude that it is possible to diagnose vasa previa by TVS combined with color Doppler in the second trimester of pregnancy. Future studies are needed to review the evidence on incidence and potential risk factors associated with vasa previa, as well as the potential protective effect of Cesarean section. This will be the most accurate way of making an informed decision on the effectiveness of routine or targeted prenatal screening for vasa previa.

ACKNOWLEDGEMENTS

LR is a PhD student at the AMC Graduate School funded by the Academic Medical Centre in Amsterdam, The Netherlands.

Chapter 2

28

REFERENCES

1. Fung TY, Lau TK. Poor perinatal outcome associated with vasa previa: is it preventable? A report of three cases and review of the literature. [Review] [43 refs]. Ultrasound in Obstetrics & Gynecology. 1998 12/1998;12(6):430-3.

2. Oyelese Y, Catanzarite V, Prefumo F, Lashley S, Schachter M, Tovbin Y, et al. Vasa previa: the impact of prenatal diagnosis on outcomes. Obstetrics & Gynecology. 2004 5/2004;103(5 Pt 1):937-42.

3. Hasegawa J, Nakamura M, Ichizuka K, Matsuoka R, Sekizawa A, Okai T. Vasa previa is not infrequent. Journal of Maternal-Fetal & Neonatal Medicine. 2012 12/2012;25(12):2795-6.


5. Kulkarni ML, Matadh PS, Ashok C, Pradeep N, Avinash T, Kulkarni AM. Absence of Wharton’s jelly around the umbilical arteries. Indian J Pediatr. 2007 8/2007;74(8):787-9.


7. Curl CW, Johnson WL. Vasa previa, antepartum diagnosis. Report of a case. Obstetrics & Gynecology. 1968 3/1968;31(3):328-30.


9. Oyelese KO, Turner M, Lees C, Campbell S. Vasa previa: an avoidable obstetric tragedy. [Review] [58 refs]. Obstetrical & Gynecological Survey. 1999 2/1999;54(2):138-45.


11. Baulies S, Maiz N, Munoz A, Torrents M, Echevarria M, Serra B. Prenatal ultrasound diagnosis of vasa praevia and analysis of risk factors. Prenatal Diagnosis. 2007 7/2007;27(7):595-9.

12. Hasegawa J, Nakamura M, Sekizawa A, Matsuoka R, Ichizuka K, Okai T. Prediction of risk for vasa previa at 9-13 weeks’ gestation. Journal of Obstetrics & Gynaecology Research. 2011 10/2011;37(10):1346-51.

13. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Annals of internal medicine. 2011 Oct 18;155(8):529-36.

14. Rebarber A, Dolin C, Fox NS, Klauser CK, Saltzman DH, Roman AS. Natural history of vasa previa across gestation using a screening protocol. Journal of Ultrasound in Medicine. 2014 1/2014;33:(1):141-7.

15. Golic M, Hinkson L, Bamberg C, Rodekamp E, Brauer M, Sarioglu N, et al. Vasa Praevia: Risk-Adapted Modification of the Conventional Management - a Retrospective Study. Ultraschall Med. 2013 8/2013;34(4):368-76.

16. Cipriano LE, Barth WH, Jr., Zaric GS. The cost-effectiveness of targeted or universal screening for vasa praevia at 18-20 weeks of gestation in Ontario. BJOG: An International Journal of Obstetrics & Gynaecology. 2010 8/2010;117(9):1108-18.


18. Derbala Y, Grochal F, Jeanty P. Vasa previa. Journal of Prenatal Medicine. 2007 1/2007;1(1):2-13.19. Rao KP, Belogolovkin V, Yankowitz J, Spinnato JA. Abnormal placentation: evidence-based diagnosis

and management of placenta previa, placenta accreta, and vasa previa. [Review]. Obstetrical & Gynecological Survey. 2012 8/2012;67(8):503-19.


29

CHAPTER

2

20. Knowles L, Attilakos G. Management and outcome of vasa praevia in a teaching hospital over a 10 year period. Archives of Disease in Childhood: Fetal and Neonatal Edition. 2011 2011;Conference: Perinatal Medicine 2011 Harrogate United Kingdom. Conference Start: 20110615 Conference End: 20110617. Conference Publication::Fa83.

21. Weintraub AY, Gutvirtz G, Sergienko R, Sheiner E. Vasa-previa: A critical analysis of risk factors and perinatal outcomes of 237 cases. American Journal of Obstetrics and Gynecology. 2012 2012;Conference: 32nd Annual Meeting of the Society for Maternal-Fetal Medicine: The Pregnancy Meeting Dallas, TX United States. Conference Start: 20120206 Conference End: 20120211. Conference Publication::S63.

22. Romero V, Perni U, Joshi D, Mozurkewich E, Treadwell, M.C. Neonatal outcomes after prenatally diagnosed vasa previa. A case series. Reproductive Sciences. 2011 2011;Conference: 58th Annual Scientific Meeting of the Society for Gynecologic Investigation, SGI 2011 Miami Beach, FL United States. Conference Start: 20110316 Conference End: 20110319. Conference Publication::183A.

23. Bronsteen R, Whitten A, Balasubramanian M, Lee W, Lorenz R, Redman M, et al. Vasa previa: clinical presentations, outcomes, and implications for management. Obstetrics & Gynecology. 2013 8/2013;122(2 Pt 1):352-7.


25. Nomiyama M, Toyota Y, Kawano H. Antenatal diagnosis of velamentous umbilical cord insertion and vasa previa with color Doppler imaging. Ultrasound in Obstetrics & Gynecology. 1998 12/1998;12(6):426-9.

26. Lijmer JG, Mol BW, Heisterkamp S, Bonsel GJ, Prins MH, van der Meulen JH, et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA. 1999 9/15/1999;282(11):1061-6.

27. Kanda E, Matsuda Y, Kamitomo M, Maeda T, Mihara K, Hatae M. Prenatal diagnosis and management of vasa previa: a 6-year review. Journal of Obstetrics & Gynaecology Research. 2011 10/2011;37(10):1391-6.


29. Hasegawa J, Farina A, Nakamura M, Matsuoka R, Ichizuka K, Sekizawa A, et al. Analysis of the ultrasonographic findings predictive of vasa previa. Prenatal Diagnosis. 2010 12/2010;30(12-13):1121-5.

30. Hasegawa J, Matsuoka R, Ichizuka K, Otsuki K, Sekizawa A, Farina A, et al. Cord insertion into the lower third of the uterus in the first trimester is associated with placental and umbilical cord abnormalities. Ultrasound in Obstetrics and Gynecology. 2006 2006;28:183-6.

31. Hasegawa J, Sekizawa A, Farina A, Nakamura M, Matsuoka R, Ichizuka K, et al. Location of the placenta or the umbilical cord insertion site in the lowest uterine segment is associated with low maternal blood pressure. BJOG: An International Journal of Obstetrics & Gynaecology. 2011 11/2011;118(12):1464-9.

32. Sepulveda W, Rojas I, Robert JA, Schnapp C, Alcalde JL. Prenatal detection of velamentous insertion of the umbilical cord: a prospective color Doppler ultrasound study. Ultrasound in Obstetrics & Gynecology. 2003 6/2003;21(6):564-9.

33. Leeflang MM. Systematic reviews and meta-analyses of diagnostic test accuracy. Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases. 2014 Feb;20(2):105-13.

34. Ventolini G, Samlowski R, Hood DL. Placental findings in low-risk, singleton, term pregnancies after uncomplicated deliveries. American Journal of Perinatology. 2004 8/2004;21(6):325-8.

35. GJ K. De pathologische zwangerschap. In: BL H, editor. De voortplanting van de mens - Leerboek voor obstetrie en gynaecologie; 1973. p. 613-5.

Chapter 2

30

36. Macones GA, Cahill A, Stamilio DM, Odibo AO. The efficacy of early amniotomy in nulliparous labor induction: a randomized controlled trial. AmJ ObstetGynecol. 2012 11/2012;207(5):403-5.

37. Jozwiak M, Oude RK, Benthem M, van BE, Dijksterhuis MG, de Graaf IM, et al. Foley catheter versus vaginal prostaglandin E2 gel for induction of labour at term (PROBAAT trial): an open-label, randomised controlled trial. Lancet. 2011 12/17/2011;378(9809):2095-103.


31

CHAPTER

2

APPENDIX

Database(s): Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations and Ovid MEDLINE(R) 1946 to Present Search Strategy: 2014-02-10

# Searches Results

1 vasa previa/ 262 (vasa* adj3 pr?evia*).tw,ot,kw. 1853 (velamentous adj2 cord insert*).tw,ot,kw. 704 (cord insert* adj3 (type*1 or kind*1)).tw,ot,kw. 65 (abnormal adj2 cord insert*).tw,ot,kw. 256 ((low or lower or os or mangrove) adj5 cord insert*).tw,ot,kw. 207 ((bilobed or succenturiate or VCI) and placenta*).tw,ot,kw. 418 (succenturiate adj3 lobe*).tw,ot,kw. 209 or/1-8 [I] 29010 Umbilical Cord/ and exp placenta diseases/ 34511 ((imaging or ultraso* or sonogra* or echo* or doppler) adj5 cord insertion*).tw,ot,kw. 2312 Umbilical Cord/ab and placenta/ 6913 (velamentous adj4 insert*).tw,ot,kw. 16314 or/11-13 22515 exp Diagnostic Imaging/ 167681816 us.fs. 19258817 (ultraso* or imaging).jw. 7724318 (imaging or ultraso* or sonogra* or echo* or doppler).tw,ot,kw. 80185019 or/15-18 [ULTRASOUND] 204492320 Mass Screening/ or Prenatal Diagnosis/ 10861121 exp “sensitivity and specificity”/ or exp “reproducibility of results”/ 56925722 (screen* or diagnos* or sensitivity or detect* or surveillance or routine or accurac* or (area

under adj5 curve) or AUC or receiver operat* or ROC).tw,ot,kw.3851200

23 or/20-22 [DIAGNOSIS] 411351124 19 or 23 [ULTRASOUND & DIAGNOSIS] 542440625 10 and 19 [II] 5326 14 and 24 [III] 11627 9 or 25 or 26 [I II III] 35328 (Animals/ or dolphin*.ti.) not Humans/ 377899629 27 not 28 [I II III -humans] 35230 “16858739”.an. 131 (“17497747” or “23969805” or “11529988” or “22839499” or “21599804” or “9918092” or

“10725492” or “11169317” or “20091699” or “22926275”).an. [DIAGN]10

32 limit 9 to yr=”2013 -Current” 2033 9 not 28 [vasa previa laden] 28934 29 not 9 6335 31 and 29 1036 31 and 24 1037 34 not 28 63

Chapter 2

32

exp “sensitivity and specificity”/ NT= predictive value of tests + ROCDatabase(s): Embase Classic+Embase 1947 to 2014 February 07 Search Strategy: 2014-02-10

# Searches Results

1 vasa previa/ 1582 (vasa* adj3 pr?evia*).tw,ot,kw. 2753 (velamentous adj2 cord insert*).tw,ot,kw. 1244 (cord insert* adj3 (type*1 or kind*1)).tw,ot,kw. 135 (abnormal adj2 cord insert*).tw,ot,kw. 506 ((low or lower or os or mangrove) adj5 cord insert*).tw,ot,kw. 217 ((bilobed or succenturiate) and placenta*).tw,ot,kw. 618 (succenturiate adj3 lobe*).tw,ot,kw. 329 or/1-8 [I] 47310 ((imaging or ultraso* or sonogra* or echo* or doppler) adj5 cord insertion*).tw,ot,kw. 3911 (placenta disorder/ or placenta accreta/ or placenta previa/) and *umbilical cord/ 7412 (velamentous adj4 insert*).tw,ot,kw. 28413 or/10-12 36414 exp flow measurement/ or exp echography/ or exp radiodiagnosis/ or ultrasound/ or diagnostic

imaging/2412166

15 (ultraso* or imaging).jw. 10793416 (imaging or ultraso* or sonogra* or echo* or doppler).tw,ot,kw. 112356817 mass screening/ or prenatal screening/ or screening/ or screening test/ or prenatal diagnosis/ 25468918 “sensitivity and specificity”/ or predictive value/ or diagnostic approach route/ or diagnostic

procedure/ or diagnostic accuracy/ or diagnostic value/ or receiver operating characteristic/561763

19 (screen* or diagnos* or sensitivity or detect* or surveillance or routine or accurac* or (area under adj5 curve) or AUC or receiver operat* or ROC).tw,ot,kw.

5302118

20 or/14-19 710755021 13 and 20 [II] 20622 9 or 21 [I II] 53223 (animal/ or animal experiment/ or animal model/ or nonhuman/ or dolphin*.ti.) not human/ 512830724 22 not 23 52825 [XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX] 026 9 not 23 47127 21 not 23 20428 27 not 26 [laden] 57

CHAPTER 3

Incidence of and risk indicators for vasa praevia:

a systematic review

Ruiter L Kok N

Limpens J Derks JB

de Graaf IMMol BWJ

Pajkrt E

British Journal of Obstetrics and Gynaecology. 2016 Jul;123(8):1278-87

Chapter 3

36

ABSTRACT

Background Vasa praevia (VP) is a rare phenomenon that is assumed to increase the risk of severe complications, including fetal death. Critical data on its incidence are lacking, so there is no rational basis for prenatal screening.

Objectives: To review the literature on the incidence and risk indicators for VP.

Search Strategy: We searched OVID MEDLINE, OVID EMBASE, the Cochrane Library and PubMed for case-control and cohort studies on incidence and risk indicators for VP.

Selection criteria: Two reviewers selected studies and scored their methodological quality.

Data collection and analysis: We calculated the mean incidence of VP. We constructed 2x2 tables cross classifying potential risk indicators against the incidence of VP to calculate common odds ratios and 95% confidence intervals, using the Mantel-Haenszel method.

Main results We included 13 studies (two prospective cohort studies, ten retrospective cohort studies and one case-control study) reporting on 569 410 patients with 325 cases of VP. Based on ten included cohort studies providing information on the incidence, the mean incidence of VP was 0.60 per 1000. We identified five different risk indicators and markers for VP: second-trimester placenta praevia, conception by assisted reproductive technologies, a bilobed or succenturiate placenta, umbilical cord insertion in the lower third part of the uterus at first-trimester ultrasound and velamentous cord insertion. Almost 83% of the cases of VP had one or more risk indicators.

Conclusions In view of the low incidence, screening for VP in an unselected population is not advised. Targeted screening of women with one or more risk indicators as a part of routine mid-gestation scanning should be considered.

Incidence of and risk indicators for vasa praevia: a systematic review

37

CHAPTER

3

INTRODUCTION

Vasa praevia (VP) is a condition in which fetal blood vessels overlie the cervical internal os, unsupported by either the umbilical cord or placental tissue. Compression of these vessels by the presenting part may lead to fetal heart decelerations and bradycardia on the cardiotocography.1, 2 Theoretically, in case of rupture of membranes, VP can rupture simultaneously, potentially causing fetal blood loss with serious neonatal morbidity or death.3

The accuracy of ultrasound in the prenatal diagnosis of VP is good with a sensitivity of 100% and specificity of 99.0 – 99.8%, when performed transvaginally with colour Doppler.4 It has been hypothesized that prenatal recognition of VP enables elective delivery of the fetus by Cesarean section, so preventing potential rupture of the low-lying vessel(s) and subsequent fetal demise or neonatal morbidity.5 Currently, the majority of pregnant women undergo several scans during pregnancy. It has been proposed to include a standard evaluation of the umbilical cord insertion site in the second-trimester scan in an attempt to increase the prenatal detection of VP. 3, 6, 7 However, at present, targeted transvaginal ultrasound screening with colour Doppler has not been routinely included in prenatal care, due to a lack of critical data on the incidence and the efficacy of screening. For the purpose of potential screening for VP it might be helpful to identify women at high risk by establishing identifiable risk indicators for VP. We performed a meta-analysis to investigate the incidence and risk indicators of VP and to define the population at risk for the purpose of a potential future screening programme.

METHODS

Identification of studiesThe research protocol was published in the International prospective register of systematic reviews (Prospero).8 A medical librarian (JL) performed a comprehensive search of MEDLINE (OVID, from 1948), EMBASE (OVID, from 1947), the Cochrane Library including the Cochrane Central Register of Controlled Trials (CENTRAL, from inception), PubMed (the publisher[sb]-fraction which contains publications ahead of print, not yet included in OVID MEDLINE) and ongoing Trial registers (http://clinicaltrials.gov/) between inception and February 2014. There were no language restrictions and animal studies were excluded. The search strategies consisted of Subject Headings (MeSH, SH) and words in title and abstract for VP or its synonyms (vasa praevia I). In addition we broadly searched for diagnostic imaging and screening of abnormal umbilical cord location and the subject of risk or association (vasa praevia II and III, see Supplementary material, Appendix S1, for the entire MEDLINE search). For each database the search strategy was refined using an iterative process through incorporation

Chapter 3

38

of new search terms when new relevant citations were identified, i.e. by checking reference lists and citing articles using ISI Web of Science. The bibliographic records retrieved were downloaded and imported into REFERENCE MANAGER® (Thomson Reuters, Carlsbad, CA, USA) software (version 12.0) to remove duplicates, and to store and analyse the search results.

Quality assessment The included studies were systematically scored on their methodological quality, based on predetermined key features using the Newcastle-Ottawa Quality Assessment Scale (NOS). Key features were cohort study design, consecutive patient enrolment and prospective study design. Additional key features for studies on potential risk indicators were a representative exposed cohort and non-exposed cohort from the same population, ascertainment of exposure and outcome by the investigator or from the medical record, absence of the outcome at start of the study period and completeness of follow-up. For case-control studies similar items were rated for the cases and controls. Two reviewers independently scored all studies on the individual items using the Newcastle-Ottawa Quality Assessment Scale.9

Study selection and data extractionStudies were eligible if they were cohort or case-control studies. A study was included in this review if (1) the study population included subjects with a confirmed or assumed diagnosis of VP (tubular structures or vessels from fetal origin visualized over the internal cervical os on ultrasound); (2) a control group without VP (or non-cases) was available; (3) at least one potential risk indicator was measured in all subjects. Case-control studies were also included when VP was the outcome of interest and exposed subjects (e.g. pregnancies with placenta praevia or by assisted reproductive technologies) were compared to their unexposed (e.g. pregnancies with normal placentation or natural conception) controls. Case reports, letters to the editor and conference abstracts were excluded. Disagreements about in- or exclusions of full text articles were resolved by discussion. Study selection was performed in a three stage process. First, two reviewers (LR, NK) scrutinized titles and abstracts of all references for reporting on VP and one or more potential risk indicators. The two reviewers then examined all references that were selected by at least one reviewer. Separate final inclusion and exclusion decisions were made for both incidence rate and potential risk indicators after duplicate examination of the full text versions. Data were extracted by one reviewer (LR) with the use of a predesigned data extraction form. Study characteristics that were summarised on the form included: country of investigation, period of data collection, total cohort size, number of cases of VP during the study period and study population. In the absence of absolute numbers on cases of VP these were calculated when possible from given percentages or proportions.10 For studies with additional data on potential risk indicators, their prevalence in the subjects and controls in


39

CHAPTER

3

cohort studies and in the cases and controls in case-control studies was extracted. Crude odds ratios (OR) had to be available either directly or, in case the paper reported adjusted odds ratios, had to be derivable from the original data supplied. If an estimate of the odds ratio for an association with a risk indicator was reported together with its precision, these numbers were used. A second reviewer (NK) examined the extracted data on accuracy and completeness. Disagreements were solved by discussion.

Statistical analysisCalculation of the incidence of VP (defined as the proportion of women who suffered VP during the study period) was based on cohort studies only. To allow for study size, the mean incidence rate of VP was calculated from the original demographic data (denominator) and the number of cases included (numerator). For each included study with information on potential risk indicators 2 x 2 tables were constructed cross classifying the potential risk indicator against the incidence of VP. Quantitative analysis was conducted using REVIEW

MANAGER (REVMAN) software version 5.3. The effect of a potential risk indicator on the incidence of VP was expressed as odds ratio with a 95% confidence interval (95% CI) and calculated using the Mantel-Haenszel method. To assess the extent of heterogeneity the magnitude of the value of I2 was used.11 When heterogeneity was present we used a random-effect model to calculate a common odds ratio and 95% confidence interval. When heterogeneity was rejected, we used a fixed-effect model. There was no adjustment for potential confounders in the meta-analysis. In addition, we calculated the number needed to screen to detect one case of VP for the individual risk indicators.

RESULTS

Figure 1 summarises the results of the literature search. The search resulted in 572 unique articles, of which 553 articles were excluded after reading title and abstract. Articles that were excluded were mainly case reports or studies that reported on umbilical cord anomalies other than VP. Nineteen studies were selected for full text reading. One study had insufficient data to estimate either the incidence rate or to construct 2 x 2 tables.12 One cohort was described twice and therefore one study was excluded.7, 13 Two studies were (conference) abstracts of articles not yet published, one author was reached but no more data were obtained.14, 15 Two studies only reported on associations between velamentous cord insertion and low serum pregnancy-associated plasma protein A or cord insertion in the lower part of the uterus in the first trimester; associations with VP itself were not investigated and therefore these studies were excluded.16, 17 Finally, we included a total of 13 unique studies, ten studies reported on the incidence rate and eight studies reported on potential risk indicators.

Chapter 3

40

Figure 1. PRISMA Flowchart

Table 1 gives an overview of the included studies and their characteristics. The total number of patients included was 569 410 including 325 cases of VP.

Quality assessmentFigure 2 summarises the results of the quality assessment of the included studies. All but one study were cohort studies. Data collection was prospective in two of 13 studies (15%), all studies performed consecutive patient enrolment. For the studies on potential risk


41

CHAPTER

3

indicators seven of eight studies (88%) reported the use of a representative exposed cohort. All studies drew the non-exposed controls from the same population as the exposed cases. Ascertainment of exposure was in four studies (50%) by the investigator or by medical record, four studies did not report ascertainment of exposure. All eight studies assessed the outcome by medical record. Furthermore, follow-up was long enough in all studies for the outcome to occur. All studies provided effect measures with confidence intervals.

Figure 2. Quality assessment

Assessment of heterogeneity and quantitative analysis Heterogeneity among studies was assessed per potential risk indicator. The results of these test demonstrated substantial heterogeneity for two risk indicators (I2 = 66% and 77%) and moderate heterogeneity for one risk indicator (I2 = 29%), consequently the random effect method was used. For the remaining three risk indicators the I2 demonstrated no heterogeneity and so the fixed effect method was used.

IncidenceFrom ten included cohort studies there was information on the incidence of VP, which varied from 0.17 to 2.2 per 1000 pregnancies (table 1). A total of 496 196 women were included in these studies of which 299 women had VP, the mean incidence of VP was 0.60 per 1000 pregnancies.

Chapter 3

42

Table 1. Characteristics of the included studies

Author, year and country of publication

Type of study Risk factors studied* Study population Incidence (per 1000 pregnancies)

Baulies, 2007 (Spain)6 Retrospective cohort

Second trimester placenta previa, bilobed placenta, ART (not reported)

All deliveries between 2000 and 2005 (n=12 063, 9 cases of VP)

0.75

Catanzarite, 2001 (USA)23

Prospective cohort

No risk factors studied All patients with ultrasound between 1991 and 1998 (n=33 208, 10 cases of VP)

0.3

Eddleman, 1992 (USA)20 Retrospective cohort

Second trimester placenta previa and VCI (not adjusted)

All deliveries between 1985 and 1988 (n=15 942, 3 cases of VP)

0.19

Francois, 2003 (Arizona)19

Case-control Second trimester placenta previa (not adjusted)

13 cases of vasa previa matched in a 1:4 ratio with controls (n=65)

N.A.

Hasegawa, 2006 (Japan)22

Prospective cohort

First trimester lower cord insertion (not adjusted)

318 patients with successful screening of lower third of uterus (n=318, 1 case of VP)

N.A.

Hasegawa, 2010 (Japan)13

Retrospective cohort

VCI, second trimester placenta previa, low umbilical cord insertion (not adjusted)

All deliveries between 2006 and 2009 (n=4532, 10 cases of VP)

2.21

Kanda, 2011 (Japan)32 Retrospective cohort

No risk factors studied All deliveries between 2002 and 2007 (n=5131, 10 cases of VP)

1.95

Lee, 2000 (USA)33 Retrospective cohort

No risk factors studied All deliveries between 1991 and 1998 (93 874, 18 cases of VP)

0.19

Rebarber, 2013 (USA)34 Retrospective cohort

No risk factors studied All patients with ultrasound between 2005 and 2012 (n=27 573, 13 cases of VP)

0.47

Rosenberg, 2011 (Israel)10


Placenta previa (not adjusted)

All singleton deliveries between 1988 - 2009 (n=185 476, 204 cases of VP)

1.1

Schachter, 2002 (Israel)18


ART (not reported) All deliveries between 1987 and 2001 (n=72 818, 12 cases of VP)

N.A.

Smorgick, 2009 (Israel)35 Retrospective cohort

No risk factors studied All deliveries between 1988 and 2007 (n=110 684, 19 cases of VP)

0.17

Suzuki, 2008 (Japan)21 Retrospective cohort

Succenturiate placenta (not adjusted)

All deliveries between 2002 and 2005 (n=7713, 3 cases of VP)

0.39

*in case of adjustment; adjusted for


43

CHAPTER

3

Risk indicatorsAmong the studies on risk indicators, at least 83% of the VP cases had one or more risk indicators except for one study. In this study only one indicator was investigated and 63% of the VP cases had this risk indicator. 18 All but one of the studies reported results from univariate analyses on associations between potential risk indicators and VP; as a consequence, unless stated otherwise, these numbers were used.6

Second trimester placenta praeviaFour studies investigated the association between placenta praevia in the second trimester and VP; one case-control study and three cohort studies with a total of 202 060 women and 236 cases of VP.6, 10, 13, 19 For women with a placenta praevia in the second trimester the common odds ratio for VP was 19 (95% CI 6.1-58, heterogeneity I2=66%) compared to women with a normal placental localisation. (Figure 3A). One study performed a multivariate analysis in which the OR remained statistically significant after correction for potential confounders; OR 23 (95% CI 5.6–93.8), however the potential confounders that were adjusted for in this study were not reported and were not obtained after contact with the authors.6 Based on the raw data, the number needed to screen for women with a second trimester placenta praevia was 63 to detect one case of VP compared to women without a second trimester placenta praevia.

Velamentous cord insertionTwo studies reported on the association between velamentous cord insertion and VP.13, 20 Velamentous cord insertion was defined as an umbilical cord insertion outside the chorionic plate. Women with velamentous insertion of the umbilical cord had an increased risk as compared to women with a normal placental cord insertion (common odds ratio 672; 95% CI 112–4034, heterogeneity I2=0%) (Figure 3B). For women with a velamentous cord insertion, the number needed to screen to find one case of VP was 13.

Assisted reproductive technologies (ART)Two studies reported on the association between ART (in vitro fertilization and intracytoplasmic sperm injection) and VP. 6, 18 Women who conceived by ART had an increased risk compared with women who conceived spontaneously (common OR 19; 95% CI 6.6–54, heterogeneity I2=29%) (Figure 3C). The number needed to screen for women conceived by ART to detect one case of VP was 260.

Bilobed placentaTwo studies reported on the association between placental morphological abnormalities (e.g. bilobed and succenturiate placenta) and VP.6, 21 Women with a bilobed or succenturiate

Chapter 3

44

Figure 3A-F. Results of meta-analysis

In order of appearance: second trimester placenta previa, velamentous cord insertion, ART, placental morphological abnormalities, lower umbilical cord insertion, multiple pregnancy.


45

CHAPTER

3

placenta had an increased risk compared with women with a normally developed placenta (common OR 71; 95% CI 14–349, heterogeneity I2=0%) (Figure 3D). The number needed to screen for women with a bilobed or succenturiate placenta to detect one case of VP was 37.

Cord insertion in the lower third of the uterus at first-trimester ultrasoundTwo studies investigated characteristics of first-trimester ultrasound scans. In these studies, the uterus was divided in three equal parts; upper, middle and lower third. A significant association was found between insertion of the umbilical cord in the lowest third of the uterus and VP.13, 22 Women with a low umbilical cord insertion had an increased risk compared with women with an umbilical cord insertion in the middle or upper third part of the uterus (common OR 280; 95% CI 1.5–51 547, heterogeneity I2=86%) (Figure 3E). In case of detection of a lower cord insertion on first trimester ultrasound, the number needed to screen to eventually find one case of VP was 6.1.

Multiple gestationOne case-control and two cohort studies reported on the association between multiple gestation and VP.6, 13, 19 The common OR for VP was 2.66 (95% CI 0.80 – 8.8, heterogeneity I2=3%) for women with a multiple gestation compared with a singleton gestation (Figure 3F).

DISCUSSION

Main findingsOur systematic review showed a mean incidence of VP of 0.60 per 1000. Placenta praevia in the second trimester, velamentous cord insertion, conception by ART, placental morphological anomalies and lower umbilical cord insertion at first trimester ultrasound were associated with an increased risk of VP.

Strengths and limitationsThe importance of our study is that based on a systematic review of the literature, several clinical factors are associated with VP. Evidence for these factors so far mainly lies in small case series and case reports so we performed a thorough search strategy without any language restrictions. We made an effort to find and screen all the available articles on VP. For the purpose of a meta-analysis cohort studies are most valuable since there is a well-defined population where cases and non-cases stem from the same population. Cohort studies are therefore likely to yield the most robust findings. In this systematic review, 12 out of 13 studies were cohort studies. Furthermore, the included studies showed consistency on the different risk indicators, all studies found similar clinical factors associated with VP. Consistent results were found in both smaller and larger studies.

Chapter 3

46

A limitation of our study is the absence of access to individual patient data which makes adjustment for potential confounders impossible. For real consideration whether the previously described risk indicators are independently associated with VP, individual patient data or large longitudinal cohort studies are needed. The possibility to adjust for confounders will allow us to develop risk adjusted prognostic models for VP, but it is difficult to achieve in, for example, the Netherlands because potential risk indicators and confounders such as mode of conception, maternal body mass index and presence of placenta praevia are non-obligatory fields in our birth register. Another limitation is the overall methodological quality of the analyzed studies, this was low and publication bias is a significant problem. The unavailability of several publications on VP more than two years after publication of a conference-abstract only supports this. Although we tried to obtain as much data as possible by contacting all authors of incomplete data, this will remain a limitation in future studies. In addition to the risks of bias, we also identified some other features that may limit the conclusions of our study. First, the small number of exposed subjects and the small number of events in most of the included studies limit their statistical power, also partially revealed by the wide confidence intervals of some effect measures. Furthermore we found a relatively large difference in incidence of VP between the prospective and retrospective cohort studies. This might be explained by the fact that prospective studies focus more on detecting VP which can lead to a higher incidence. This will almost certainly be accompanied by more false-positive cases as well, since the validity of postpartum confirmation of VP is limited.4 Hence, defining the true incidence will also remain limited. In contrast, retrospective studies are known to have limitations from incompleteness of patient data; some incidences of VP and the presence of potential clinical risk factors might not have been accurately reported in the past. However, since VP is a rather rare complication, a prospective study with sufficient statistical power is not feasible.

InterpretationIn this meta-analysis almost 83% of all cases of VP had one or more clinical risk indicators for VP. This rate is potentially even higher since information on presence or absence was not available for all risk factors in all individual studies. Unadjusted numbers needed to screen varied between 6 and 260; screening for VP can thus potentially be effective in a specific population of pregnant women such as women with a velamentous cord insertion or a bilobed placenta. The limited evidence on the positive predictive value of an ultrasound diagnosis of VP indicate that the majority of positively screened patients will be truly diseased, however, unfortunately exact numbers are not available, limiting the usefulness of these results.4


47

CHAPTER

3

We found velamentous cord insertion to have a remarkably strong association with VP; some studies reported that all cases of VP were found among women with a velamentous cord insertion. VP in those studies was not found in women with a normal cord insertion. Catanzarite et al. reported a case series on VP and classified two types. Type I in which there is a single placental lobe with a velamentous cord insertion over the cervix and type II in which the vessels over the cervix are connecting multiple lobes of the placenta.23 Eight out of ten cases were type I. Together with the results of this meta-analysis we might conclude that velamentous cord insertion is not a risk indicator for VP but rather a marker for VP. Based on these results one cannot rule out VP after a normal cord insertion is seen in the second trimester, but the risk becomes significantly smaller. On the contrary, once a velamentous cord insertion is seen, obstetric care givers should be aware of the possible presence of VP. Potentially, the same explanation could apply for the significant association we found between a placental morphological abnormality, such as a succenturiate or bilobed placenta, and VP. We might approach a bilobed placenta, as well, as a marker of VP and not a risk indicator. A second trimester placenta praevia that is not connected to the internal ostium in the third trimester is also a risk indicator for VP. A potential explanation might be trophotropism, a unidirectional lateral growth of the placenta in the fundal direction to ensure blood supply from a more richly vascularized area.24 Any fetal blood vessels embedded in the placenta in the cervical area will become unprotected once the placental tissue atrophies, resulting in velamentous vessels close to the cervix. Therefore, especially women with a resolved placenta praevia in the third trimester appear to be at risk for VP. The etiology behind pregnancies resulting from ART and VP has been described before. It is hypothesized that abnormal cord insertion is a result from an oblique orientation of the blastocyst at nidation.25-27

Most authors of case-reports report multiple pregnancies as a risk indicator for VP, how ever our data fail to support this finding. Potentially, the assumed association between multiple pregnancies and VP can be explained by ART. Differentiating between multiple pregnancies by ART and by spontaneous conception would be informative; however we did not have access to these data. Regarding the feasibility of the risk indicators we must allow for the fact that some indicators are easy to recognise antenatally such as ART pregnancies and a second trimester placenta praevia whereas for other indicators it is more difficult to recognise them antenatally such as velamentous cord insertion, a bilobed or succenturiate placenta and a first trimester low cord insertion. Focusing on easy to recognise indicators for VP can potentially increase the prenatal detection. Additionally, any suspicion of risk indicators legitimises referral to more experienced sonographers for further evaluation.

Chapter 3

48

CONCLUSION

In conclusion, since the majority of VP is preceded by one or more risk factors, screening for VP as part of routine mid-gestation scanning in the general low-risk population is not advised. This conclusion is supported by Cipriano et al. who demonstrated that general screening for VP is non-cost-effective.28 However, with several clinical factors being associated with VP, one could consider implementing a screening program for patients presenting with one of these factors as suggested by Derbala et al.29 Nevertheless, even then, the incidence of VP remains low. Moreover, the aforementioned risk indicators and markers such as conception by ART and a second trimester placenta praevia are relatively common. Subjecting all people with at least one risk indicator or marker to a screening initiative may not always be beneficial if one thinks in terms of uncertainty and generation of stress for patients as consequences of false-positive cases, the increasing number of Cesarean sections and the additional healthcare costs.30 These consequences should be taken into account before general introduction of a screening strategy for VP. Besides the practical issues there is a general lack of knowledge of the potential risk factors for VP. In a national questionnaire survey Ioannou and Wayne found 34% of all questioned obstetricians not sufficiently expert to name at least one risk factor for VP – implying that there is more room for improvement of outcome in VP than screening alone.31

Before the introduction of a targeted screening strategy for patients with one or more risk indicators in prenatal care, large cohort studies can be helpful to develop risk adjusted prognostic models. Furthermore a thorough decision analysis should be performed to determine the best strategy to reduce fetal mortality due to VP. Meanwhile, it seems justified to exclude VP by transvaginal ultrasound with color Doppler in women with a velamentous cord insertion or a bilobed or succenturiate placenta. Moreover, it seems justified to add a transvaginal scan with colour Doppler in the third trimester in all women diagnosed with a placenta overlying the cervix in the second trimester.

ACKNOWLEDGEMENTS

LR is a PhD student at the AMC Graduate School funded by the Academic Medical Centre in Amsterdam.


49

CHAPTER

3

REFERENCES



3. Oyelese KO, Turner M, Lees C, Campbell S. Vasa previa: an avoidable obstetric tragedy. [Review] [58 refs]. Obstetrical & Gynecological Survey. 1999 2/1999;54(2):138-45.

4. Ruiter L, Kok N, Limpens J, Derks JB, de Graaf IM, Mol BW, et al. A systematic review on the diagnostic accuracy of ultrasound in the diagnosis of vasa previa. Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology. 2014 Dec 10.




8. Prospective register of systematic reviews. http://www.crd.york.ac.uk/PROSPERO/. [cited; Available from:

9. Wells G.A SB, O’Connell D.,Peterson J.,Welch V.,Losos M.,Tugwell P,. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. 2011.

10. Rosenberg T, Pariente G, Sergienko R, Wiznitzer A, Sheiner E. Critical analysis of risk factors and outcome of placenta previa. Archives of Gynecology & Obstetrics. 2011 7/2011;284(1):47-51.

11. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in medicine. 2002 Jun 15;21(11):1539-58.

12. Daniel Y, Schreiber L, Geva E, Amit A, Pausner D, Kupferminc MJ, et al. Do placentae of term singleton pregnancies obtained by assisted reproductive technologies differ from those of spontaneously conceived pregnancies? Human Reproduction. 1999 4/1999;14(4):1107-10.


14. Matsuoka R, Hasegawa J, Aoki H, Koide K, Ichizuka K, Sekizawa A, et al. 15 cases of vasa previa. Journal of Perinatal Medicine. 2011 2011;Conference: 10th World Congress of Perinatal Medicine 2011 Punta del Este Uruguay. Conference Start: 20111108 Conference End: 20111111. Conference Publication:.

15. Weintraub AY, Gutvirtz G, Sergienko R, Sheiner E. Vasa-previa: A critical analysis of risk factors and perinatal outcomes of 237 cases. American Journal of Obstetrics and Gynecology. 2012 2012;Conference: 32nd Annual Meeting of the Society for Maternal-Fetal Medicine: The Pregnancy Meeting Dallas, TX United States. Conference Start: 20120206 Conference End: 20120211. Conference Publication::S63.

16. Hasegawa J, Farina A, Simonazzi G, Bisulli M, Puccetti C, Pilu G, et al. Umbilical cord insertion into the lower segment of the uterus at 11 to 13 weeks’ gestation is associated with maternal serum PAPP-A. Prenatal Diagnosis. 2011 2011;31:434-8.

17. Raisanen S, Georgiadis L, Harju M, Keski-Nisula L, Heinonen S. Risk factors and adverse pregnancy outcomes among births affected by velamentous umbilical cord insertion: a retrospective population-based register study. European Journal of Obstetrics, Gynecology, & Reproductive Biology. 2012 12/2012;165(2):231-4.

Chapter 3

50

18. Schachter M, Tovbin Y, Arieli S, Friedler S, Ron-El R, Sherman D. In vitro fertilization is a risk factor for vasa previa. Fertility & Sterility. 2002 9/2002;78(3):642-3.

19. Francois K, Mayer S, Harris C, Perlow JH. Association of vasa previa at delivery with a history of second-trimester placenta previa. Journal of Reproductive Medicine. 2003 10/2003;48(10):771-4.


21. Suzuki S, Igarashi M. Clinical significance of pregnancies with succenturiate lobes of placenta. Archives of Gynecology & Obstetrics. 2008 4/2008;277(4):299-301.

22. Hasegawa J, Matsuoka R, Ichizuka K, Otsuki K, Sekizawa A, Farina A, et al. Cord insertion into the lower third of the uterus in the first trimester is associated with placental and umbilical cord abnormalities. Ultrasound in Obstetrics and Gynecology. 2006 2006;28:183-6.


24. Monie IW. Velamentous Insertion of the Cord in Early Pregnancy. Am J Obstet Gynecol. 1965 Sep 15;93:276-81.

25. Al-Khaduri M, Kadoch IJ, Couturier B, Dube J, Lapensee L, Bissonnette F. Vasa praevia after IVF: should there be guidelines? Report of two cases and literature review. Reproductive Biomedicine Online. 2007 3/2007;14(3):372-4.

26. Englert Y, Imbert MC, Van RE, Belaisch J, Segal L, Feichtinger W, et al. Morphological anomalies in the placentae of IVF pregnancies: preliminary report of a multicentric study. Human Reproduction. 1987 2/1987;2(2):155-7.

27. Gavriil P, Jauniaux E, Leroy F. Pathologic examination of placentas from singleton and twin pregnancies obtained after in vitro fertilization and embryo transfer. Pediatric pathology / affiliated with the International Paediatric Pathology Association. 1993 Jul-Aug;13(4):453-62.


29. Derbala Y, Grochal F, Jeanty P. Vasa previa. Journal of Prenatal Medicine. 2007 1/2007;1(1):2-13.30. Javid N, Sullivan EA, Halliday LE, Duncombe G, Homer CS. “Wrapping myself in cotton wool”:

Australian women’s experience of being diagnosed with vasa praevia. BMC Pregnancy Childbirth. 2014;14:318.

31. Ioannou C, Wayne C. Diagnosis and management of vasa previa: a questionnaire survey. Ultrasound in Obstetrics & Gynecology. 2010 2/2010;35(2):205-9.

32. Kanda E, Matsuda Y, Kamitomo M, Maeda T, Mihara K, Hatae M. Prenatal diagnosis and management of vasa previa: a 6-year review. Journal of Obstetrics & Gynaecology Research. 2011 10/2011;37(10):1391-6.


34. Rebarber A, Dolin C, Fox NS, Klauser CK, Saltzman DH, Roman AS. Natural history of vasa previa across gestation using a screening protocol. Journal of Ultrasound in Medicine. 2014 1/2014;33:(1):141-7.

35. Smorgick N, Tovbin Y, Ushakov F, Vaknin Z, Barzilay B, Herman A, et al. Is neonatal risk from Vasa previa preventable? The 20-year experience from a single medical center. Obstetrical and Gynecological Survey. 2010 2010;65:299-301.


51

CHAPTER

3

APPENDIX S1. Details of literature search ‘Incidence of and risk indicators for vasa previa’Database(s): Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations and Ovid MEDLINE(R) 1946 to Present Search Strategy: 2014-02-10

# Searches Results

1 vasa previa/ 262 (vasa* adj3 pr?evia*).tw,ot,kw. 1853 (velamentous adj2 cord insert*).tw,ot,kw. 704 (cord insert* adj3 (type*1 or kind*1)).tw,ot,kw. 65 (abnormal adj2 cord insert*).tw,ot,kw. 256 ((low or lower or os or mangrove) adj5 cord insert*).tw,ot,kw. 207 ((bilobed or succenturiate or VCI) and placenta*).tw,ot,kw. 418 (succenturiate adj3 lobe*).tw,ot,kw. 209 or/1-8 [I] 29010 Umbilical Cord/ab and placenta/ 6911 (velamentous adj4 insert*).tw,ot,kw. 16312 or/10-11 [I] 216

13 probability/ or likelihood functions/ or odds ratio/ or proportional hazards models/ or exp risk/ or exp regression analysis/ 1015914

14 case-control studies/ or exp cohort studies/ or Registries/ or exp Medical History Taking/ 1482699

15

(risk* or odds or RR or likelihood or proportional hazard* or regression or multivaria* or univaria* or logistic or predict* or predispos* or caus* or pathogenes* or cohort* or prospective* or follow-up or case-control* or retrospective* or population-based or chart* or registr*).tw,ot,kw.

4768915

16 or/13-15 542506017 12 and 16 [II] 12018 9 or 17 [I II] 32419 (Animals/ or dolphin*.ti.) not Humans/ 377899620 18 not 19 [I II humans] 32321 “16858739”.an. 1

22(“17497747” or “21599804” or “20091699” or “21564407” or “20652281” or “20872421” or “17938944” or “14619643” or “12215350” or “1590867” or “70632866” or “16858739”).an. [RISK]

11

23 9 and 22 1124 limit 9 to yr=”2013 -Current” 2025 9 not 19 28926 22 and 16 1127 17 not 9 3428 27 not 19 34

29 9 and 19 1

Chapter 3

52

Database(s): Embase Classic+Embase 1947 to 2014 February 07Search Strategy: 2014-02-10

# Searches Results

1 vasa previa/ 1582 (vasa* adj3 pr?evia*).tw,ot,kw. 2753 (velamentous adj2 cord insert*).tw,ot,kw. 1244 (cord insert* adj3 (type*1 or kind*1)).tw,ot,kw. 135 (abnormal adj2 cord insert*).tw,ot,kw. 506 ((low or lower or os or mangrove) adj5 cord insert*).tw,ot,kw. 217 ((bilobed or succenturiate) and placenta*).tw,ot,kw. 618 (succenturiate adj3 lobe*).tw,ot,kw. 329 or/1-8 [I] 47310 ((imaging or ultraso* or sonogra* or echo* or doppler) adj5 cord insertion*).tw,ot,kw. 3911 (placenta disorder/ or placenta accreta/ or placenta previa/) and *umbilical cord/ 7412 (velamentous adj4 insert*).tw,ot,kw. 28413 or/10-12 364

14probability/ or “prediction and forecasting”/ or adverse outcome/ or prediction/ or predictive value/ or risk/ or risk assessment/ or risk benefit analysis/ or risk factor/ or exp regression analysis/ or multivariate analysis/

1661054

15 exp case control study/ or cohort analysis/ or longitudinal study/ or prospective study/ or retrospective study/ or register/ or medical history/ 950817

16

(risk* or odds or RR or likelihood or proportional hazard* or regression or multivaria* or univaria* or logistic or predict* or predispos* or caus* or pathogenes* or cohort* or prospective* or follow-up or case-control* or retrospective* or population-based or chart* or registr*).tw,ot,kw.

6596711

17 or/14-16 708195918 13 and 17 [II] 20619 9 or 18 [I II] 54820 (animal/ or animal experiment/ or animal model/ or nonhuman/ or dolphin*.ti.) not human/ 512830721 19 not 20 54322 [XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX] 023 9 not 20 47124 18 not 20 20325 24 not 23 [laden] 7226 10 and 17 27

CHAPTER 4

Screening, diagnostiek en beleid bij

vasa previa – concept richtlijn

Ruiter L Derks JBPajkrt E

In preparation

Chapter 4

56

DEFINITIE

Vasa previa betreft een situatie waarin foetale bloedvaten door de vliezen verlopen, buiten de choriale plaat van de placenta, over of nabij het cervicale kanaal. Er zijn twee typen vasa previa te onderscheiden; type I waarbij er sprake is van een velamenteuze navelstrenginsertie en type II waarbij de foetale vaten meerdere lobben van een placenta bilobata of succenturiata verbinden. De foetale vaten worden in dit geval niet beschermd door de placenta danwel de Whartonse gelei uit de navelstreng en zijn daarmee kwetsbaar.

INCIDENTIE

De incidentie van vasa previa is 0.6 per 1,000 zwangerschappen, ofwel 1 op 1,667, op basis van een recente meta-analyse.1 Indien de diagnose antenataal niet wordt vastgesteld, is de perinatale sterfte circa 50%. Indien vasa previa antenataal wel wordt vastgesteld en er tijdig een electieve sectio Caesarea wordt verricht, bedraagt de perinatale sterfte <5%.2

RISICOFACTOREN

In een recente meta-analyse werden de volgende risicofactoren gevonden voor vasa previa: een velamenteuze insertie (OR 672; 95%BI 112-4033), een meerlobbige placenta (OR 72; 95%BI 15-350), een laagliggende placenta in het tweede trimester (OR 19; 95%BI 6-59) en conceptie door middel van in vitro fertilisatie (OR 19; 7-54). Een screeningsbeleid gebaseerd op zulke risicofactoren kan de numbers needed to screen (ofwel het aantal zwangeren met de betreffende risicofactor dat gescreend dient te worden om één geval van vasa previa antenataal te detecteren) reduceren. (tabel 1) In een algemene populatie zwangeren moeten 1667 vrouwen gescreend worden om één geval van vasa previa op te sporen, dit aantal kan verlaagd worden, bijvoorbeeld bij een zwangere met een velamenteuze insertie. Een insertie van de navelstreng zichtbaar in het onderste 1/3e deel van de uterus op een eerste trimester echografie is ook beschreven, echter met een zeer beperkte significantie (OR 280; 95%BI 1.5-51547). Meerlingzwangerschappen werden voorheen beschouwd als risicofactor, echter uit de bovengenoemde meta-analyse bleek geen duidelijke relatie met vasa previa te bestaan. 1

Screening, diagnostiek en beleid bij vasa previa – concept richtlijn

57

CHAPTER

4

Tabel 1. Risicofactoren voor vasa previa met numbers needed tot screen

Risicofactor Number needed to screen

Velamenteuze insertie 13Placenta bilobata / succenturiata 37Placenta previa 63Zwanger na IVF of ICSI 260Zwangere in algemene populatie 1667

Ten aanzien van placenta previa betreft het met name de laag liggende placenta in het tweede trimester die in het derde trimester is opgetrokken, mogelijk ten gevolge van unidirectionele groei van de placenta in de richting van de goed gevasculariseerde fundus van de uterus. Hierbij atrofieert het placentaire weefsel in de cervicale regio waarbij de foetale vaten van dit deel van de placenta bloot komen te liggen en er vasa previa kan optreden.3

Circa 85% van de vrouwen met vasa previa heeft een of meerdere van bovengenoemde risicofactoren. Dit percentage ligt mogelijk nog hoger, aangezien meerdere studies in de meta-analyse slechts één risicofactor onderzochten. Aangezien nagenoeg elke vasa previa getypeerd kan worden als een velamenteuze insertie (type I) danwel een placenta bilobata of succenturiata (type II) is de kans op vasa previa in geval van een normale navelstrenginsertie en een normaal gelokaliseerde eenlobbige placenta uiterst gering.

SYMPTOMEN

Vasa previa verloopt vaak asymptomatisch zolang de vliezen niet breken. Er is echter beschreven dat compressie door het voorliggende deel kan leiden tot foetale deceleraties en bradycardie op het cardiotocogram (CTG) in afwezigheid van gebroken vliezen.4-6 Op het moment dat de vliezen wel breken, kunnen de foetale vaten scheuren. Hierbij is er kans op verbloeding van het kind met potentiele asfyxie en neonatale sterfte als gevolg.

DIAGNOSTIEK

Vasa previa kan antenataal worden gediagnosticeerd met behulp van echografie. Diagnostiek kan het beste plaats vinden met een transvaginale echo en kleurendoppler in het tweede trimester van de zwangerschap. In twee prospectieve studies zijn de sensitiviteit (100%) en specificiteit (99.0 – 99.8%) van transvaginale kleurendoppler hoog.7, 8 De prenatale detectie rate varieerde tussen de 53% en 100%. 7-11 De diagnose vasa previa wordt vaker gemist

Chapter 4

58

wanneer er enkel trans abdominaal wordt gekeken of uitsluitend in het derde trimester van de zwangerschap.12 Ook onder optimale omstandigheden kan de diagnose vasa previa echter gemist worden. Wij adviseren om bij alle zwangeren de navelstrenginsertie in de placenta te beoordelen ten tijde van het structureel echoscopisch onderzoek (het SEO) en dit te documenteren (zie protocol SEO). In het geval van een velamenteuze insertie of een placenta bilobata dient een vaginale echo met kleuren Doppler te worden verricht om vasa previa uit te sluiten. In het geval van een placenta previa dient een transvaginale echoscopische evaluatie te worden verricht bij 32 weken zwangerschap ter uitsluiting van vasa previa.Aan vasa previa dient te worden gedacht wanneer er echolucente structuren zichtbaar zijn over of nabij het cervicale kanaal. In het geval van verdenking vasa previa van foetale arteriële vaten kan met pulse waved (color) Doppler relatief eenvoudig worden beoordeeld of er sprake is van een foetale hartslag en daarmee vasa previa of dat het maternale vaten betreffen. In het geval van foetale venen is dit lastiger en zal moeten worden beoordeeld of deze ontspringen uit een deel van de placenta danwel uit de navelstreng. Als dit niet het geval is, zullen het maternale vaten zijn en is er geen sprake van vasa previa. Durante partu moet aan vasa previa gedacht worden bij onverklaarde deceleraties of een bradycardie op het CTG bij staande vliezen.13 Door middel van een vaginaal toucher kan gecontroleerd worden op foetale pulsaties en kan vasa previa gediagnosticeerd worden. Ook kan een amnioscopie verricht worden om te bekijken of er foetale vaten zichtbaar zijn 14 hoewel de praktijk leert dat dit nog weinig gebeurd. Transvaginale echografie is hedendaags een veelgebruikte en snelle vorm van diagnostiek naar vasa previa, maar men moet met name in het derde trimester bedacht zijn op bemoeilijkende factoren in de beeldvorming door een ingedaald voorliggend deel, een sectio litteken, maternale habitus en vulling van de blaas. Daarnaast zullen vasa previa die bestaan uit meerdere kleine vaten tussen een meerlobbige placenta moeizamer te diagnosticeren zijn dan een velamenteuze insertie nabij het cervicale kanaal.

ScreeningUit meerdere studies is gebleken dat vrouwen met vasa previa in ca. 85% van de gevallen een of meerdere risicofactoren hadden.1, 15 Uit een kosteneffectiviteitsanalyse van Cipriano uit 2010 is gebleken dat universele screening op vasa previa niet kosteneffectief is.16 In deze studie werd aangehouden dat bij een strategie met een incremental cost-effectiveness ratio (ICER) van minder dan $20,000 per quality-adjusted life year (QALY) sterk aanbevolen wordt deze methode te implementeren.17 De ICER van universele screening naar vasa previa was $579,164 per quality-adjusted life year. Bij de strategie waarbij er een aanvullende transvaginale echografie met kleuren Doppler bij zwangeren met een velamenteuze navelstrenginsertie, een meerlobbige placenta, een placenta previa of een zwangerschap na fertiliteitsbehandeling verricht wordt, was de ICER $15,764 per QALY.


59

CHAPTER

4

Figuur 1. Flowchart beoordeling placenta en navelstrenginsertie ten tijde van het structureel echoscopisch onderzoek (het SEO)

Chapter 4

60

In de studies waarin de associatie tussen vasa previa en fertiliteitsbehandelingen werd onderzocht, is geen onderscheid gemaakt naar de lokalisatie van de placenta en kan er derhalve geen duidelijk onderscheid gemaakt worden of het verhoogde risico op basis van de fertiliteitsbehandeling zelf is, of op basis van het gegeven dat er bij zwangerschappen na fertiliteitsbehandeling vaker sprake is van een placenta previa en velamenteuze navelstrenginsertie ten gevolge van een verstoorde oriëntatie van de blastocyste. Gezien de hoge number needed to screen adviseren wij vooralsnog om bij vrouwen met een zwangerschap na fertiliteitsbehandeling – net als bij vrouwen die zwanger zijn na spontane conceptie- echografisch te beoordelen of er sprake is van een placenta previa, placenta bilobata of velamenteuze insertie en alleen in die gevallen te screenen op vasa previa.

BELEID

In afwezigheid van gerandomiseerde klinische studies zijn adviezen ten aanzien van het beleid gebaseerd op beslismodellen en consensus. Een electieve sectio Caesarea bij 34 of 35 weken zwangerschap na voorbereiding met corticosteroïden heeft de voorkeur met betrekking tot de korte- en lange termijn perinatale uitkomsten, gebaseerd op een beslisanalyse uit 2011.18 Deze studie liet echter de obstetrische voorgeschiedenis buiten beschouwing. Voor vrouwen met een eerdere vroeggeboorte <34 weken kan een vroegere electieve sectio overwogen worden. Vrouwen met een of meerdere terme partus in de voorgeschiedenis zouden bij voorkeur vanaf 35 weken gepland worden in plaats van 34 weken. Onder geen enkele omstandigheid is er voordeel te verwachten van een sectio Caesarea gepland na 37 weken zwangerschap. Begeleiding tijdens de zwangerschap dient in de tweede lijn plaats te vinden. De kans dat de vliezen breken is circa 1% tussen 26 en 34 weken en 1.5% tussen 34 en 37 weken.19 Ziekenhuisopname in het derde trimester kan derhalve overwogen worden, met als doel dat de signalen van een beginnende partus tijdig worden herkend en een sectio verricht kan worden voor het breken van de vliezen.20 Objectieve data voor de effectiviteit van ziekenhuisopname ontbreken echter. In asymptomatische zwangeren kunnen de controles derhalve ook poliklinisch plaatsvinden, eventueel met monitoring van de cervixlengte. In een retrospectieve cohort studie werden zwangeren met vasa previa poliklinisch vervolgd met instructies wanneer direct contact op te nemen. Alle neonaten werden in goede conditie geboren bij een gemiddelde zwangerschapsduur van 36 weken.9

In het geval van een symptomatische vasa previa dient een spoed sectio Caesarea te worden verricht als de neonatale levenskansen voldoende geacht worden.


61

CHAPTER

4

REFERENTIES

1. Ruiter L, Kok N, Limpens J, Derks JB, de Graaf IM, Mol B, et al. Incidence of and risk indicators for vasa praevia: a systematic review. BJOG : an international journal of obstetrics and gynaecology. 2016 Jul;123(8):1278-87.

2. Oyelese Y, Catanzarite V, Prefumo F, Lashley S, Schachter M, Tovbin Y, et al. Vasa previa: the impact of prenatal diagnosis on outcomes. Obstetrics & Gynecology. 2004 5/2004;103(5 Pt 1):937-42.

3. Monie IW. Velamentous Insertion of the Cord in Early Pregnancy. American journal of obstetrics and gynecology. 1965 Sep 15;93:276-81.


5. Curl CW, Johnson WL. Vasa previa, antepartum diagnosis. Report of a case. Obstetrics & Gynecology. 1968 3/1968;31(3):328-30.


7. Nomiyama M, Toyota Y, Kawano H. Antenatal diagnosis of velamentous umbilical cord insertion and vasa previa with color Doppler imaging. Ultrasound in Obstetrics & Gynecology. 1998 12/1998;12(6):426-9.





12. Ruiter L, Kok N, Limpens J, Derks JB, de Graaf IM, Mol BW, et al. A systematic review on the diagnostic accuracy of ultrasound in the diagnosis of vasa previa. Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology. 2014 Dec 10.

13. Kruitwagen RF, Nijhuis JG. Ruptured vasa praevia indicated by a sinusoidal fetal heart rate pattern: a case report. European journal of obstetrics, gynecology, and reproductive biology. 1991 Apr 16;39(2):147-50.

14. Barham KA. The diagnosis of a vas praevium by amnioscopy. Medical Journal of Australia. 1968 1968;2-55:398-400.

15. Bronsteen R, Whitten A, Balasubramanian M, Lee W, Lorenz R, Redman M, et al. Vasa previa: clinical presentations, outcomes, and implications for management. Obstetrics & Gynecology. 2013 8/2013;122(2 Pt 1):352-7.


17. Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations. CMAJ. 1992 Feb 15;146(4):473-81.

Chapter 4

62

18. Robinson BK, Grobman WA. Effectiveness of timing strategies for delivery of individuals with vasa previa. Obstetrics & Gynecology. 2011 3/2011;117(3):542-9.

19. (NVOG) DAoOaG. Guideline Antepartum Spontaneous Rupture of Membranes 2002 [cited; Available from:

20. Gagnon R, Morin L, Bly S, Butt K, Cargill YM, Denis N, et al. Guidelines for the management of vasa previa. J Obstet Gynaecol Can. 2009 Aug;31(8):748-60.

PART IIPLACENTA ASSOCIATED COMPLICATIONS

CHAPTER 5

Predictors for emergency Caesarean delivery

in women with placenta previa

Ruiter L Eschbach SJ

Burgers MOude Rengerink K

Van Pampus MGVan der Goes BY

Mol BWJDe Graaf IM

Pajkrt E

American Journal of Perinatology. 2016 Dec;33(14):1407-1414

Chapter 5

68

ABSTRACT

Objective The objective of this study was to identify the predictors of emergency delivery inwomen with placenta previa.

Methods This is a retrospective study of pregnancies complicated by placenta previa, scheduled for a Cesarean delivery between 2001 and 2011. Using univariable and multivariable regression predictors for emergency delivery in these women were determined. Predictive performance was assessed using receiver operating characteristic analysis and calibration plot. Internal validation was performed by bootstrap analysis.

Results Of 214 women with singleton pregnancies, 93 (43%) had an emergency Cesarean delivery, and 43 (20%) were preterm. Independent predictors for emergency delivery were history of Cesarean section (odds ratio [OR] 4.7; 95% confidence interval [CI] 1.2-12), antepartum bleeding with one (OR 7.5; 95% CI 2.5-23), two (OR 14; 95% CI 4.3-47) and three or more episodes (OR 27; 95% CI 8.3-90) as well as need for blood transfusion (OR 6.4; 95% CI 1.7-23). For emergency preterm delivery, covariates were comparable. The area under the curve was 0.832 on the original data and 0.821 on the bootstrap samples.

Conclusion Predictors for emergency delivery in women with placenta previa can be used for individualized antenatal care concerning timing of delivery and corticosteroid cover. Potentially, careful selection in women with placenta previa can result in more conservative treatment in an outpatient setting and reduction of iatrogenic preterm delivery.

Predictors for emergency Caesarean delivery in women with placenta previa

69

CHAPTER

5

INTRODUCTION

When scheduling the delivery in women with placenta previa, maternal and neonatal risks need to be balanced. To allow optimal fetal maturation, Cesarean delivery is generally recommended and scheduled after 37 to 38 weeks’ gestation by international guidelines.1 However, up to 40% of women with placenta previa deliver before the scheduled date of Cesarean delivery in an emergency or preterm setting2, 3 and this is associated with peripartum hemorrhage, blood transfusion and hysterectomy, 4-7 as well as neonatal anemia, prematurity, and mortality.2, 8, 9

It is important to distinguish between women at high and low risk for emergency delivery. Women at high risk could receive corticosteroids and might have their Cesarean delivery scheduled earlier, whereas women at low risk could be monitored without further intervention until 38 weeks.

Currently, there are no data that predict the risk of emergency delivery in women with placenta previa. In this study we aim to identify factors that predict the risk of emergency delivery in order to select cases in which additional prenatal care and adapted timing of delivery can be considered and in which patients a more conservative treatment is justified.

METHODS

Setting and populationWe identified women with a singleton pregnancy beyond 24 weeks scheduled for Cesarean delivery because of placenta previa between January 2001 and December 2011 by reviewing all medical charts in one academic hospital and two teaching hospitals in the Netherlands. Pregnancies complicated by preeclampsia or hemolysis, elevated liver enzymes and low platelets (HELLP) syndrome were excluded, since emergency delivery might be indicated for other reasons than the location of the placenta. Gestational age was based on last menstrual period and ultrasound measurements in the first trimester.

Classification of placenta previa is traditionally based on the relationship of the placenta to the internal cervical os and is reported as complete or partial. The term marginal placenta previa is used for those cases with placental tissue within 20 millimetres from the internal os. The diagnosis placenta previa is generally made during the transabdominal second trimester fetal anomalies scan, performed by certified sonographers. If placenta previa is seen, placental

Chapter 5

70

localization will be evaluated by transvaginal ultrasound in the third trimester. In our study the most recent ultrasound was used.

Clinical management of placenta previaPatients were admitted if vaginal bleeding was a reason for admission. If placenta previa was still present in the early third trimester, a Cesarean delivery was scheduled from 37 weeks’ gestation according to national guidelines.10

An emergency delivery was performed in cases of vaginal bleeding, (premature) onset of labor or suspected fetal distress. Women with placenta previa presenting with symptoms of labor before 34 weeks were treated with corticosteroid cover for fetal lung maturation and tocolytics if indicated. Emergency vaginal delivery was only accepted in patients already in advanced stage of labor at first evaluation.

Outcome and candidate predictorsThe outcome was emergency delivery, that is, any delivery in emergent or semi-emergent setting before the scheduled date of Cesarean delivery documented in the medical chart. Emergency delivery was stratified in three groups; major or recurrent hemorrhage without signs of labor (‘hemorrhage’), premature onset of labor described as contractions and/or ruptured membranes with or without hemorrhage (‘labor’), and suspected fetal distress defined as nonreassuring fetal patterns on cardiotocography (‘NRFS’).

Candidate predictors were identified from literature and based on clinical reasoning. We extracted maternal characteristics including age, ethnicity, smoking and body mass index (BMI) and pregnancy characteristics including obstetric history, type of placenta previa (complete, partial, marginal based on ultrasound examination), and total amount of blood loss from the medical charts. Severity of antepartum bleeding was defined by the number of bleeding episodes, duration of hospital admission (more or less than one week), use of medication and need for blood transfusion to compensate antepartum bleeding. The criteria for blood transfusion were hemodynamic instability and the pursuit of hemoglobin level > 6.0mmol/L. An episode of vaginal bleeding was defined as an admission or a presentation at the emergency room. AnalysisA univariable preselection process was used to select variables with a p-value ≤ 0.20. A multivariable model was constructed to identify independent predictors for emergency delivery in placenta previa using backward stepwise elimination. Variables with a p-value < 0.05 remained in the final model. When more than one variable for severity of bleeding


71

CHAPTER

5

was significant, the variable with the most predictive value was used in the final model. Associations were expressed as odds ratios (ORs) with 95% confidence intervals (CIs). To test for differences between variables, chi-square and t-test were used where appropriate.

To evaluate the performance of the model, a calibration plot was constructed using five nearly equal sized groups based on predicted chance of emergency delivery. The area under the receiver operating characteristic (ROC) curve was studied, an area under the curve (AUC) of 1 means the model predicts perfectly in all cases, whereas an AUC of 0.5 means the model does not predict emergency delivery at all. Interval validation of the model was performed by bootstrapping to correct for optimistic prediction. We generated 1,000 bootstrap samples with replacement and evaluated the difference in AUC on the bootstrap sample and the original sample.11

Kaplan-Meier curves were constructed for women delivering before and on the scheduled date as well as for the three indications for emergency delivery. In addition, we compared the predictive factors between women that delivered early preterm before 34 weeks’ gestation and women that delivered late preterm between 34 and 37 weeks’ gestation.

RESULTS

A total of 54,794 women delivered in the three hospitals in the study period, and 243 (0.44%) women were diagnosed with placenta previa. We excluded 29 women; eight women had a twin pregnancy, six had preeclampsia/HELLP, three had immature intra uterine fetal demise (IUFD), one had multi organ failure leading to premature delivery and 11 had no scheduled Cesarean delivery because of placenta previa marginalis with intended vaginal delivery. (Figure 1).

Of 214 women included, 99% (212/214) underwent Cesarean delivery. The gestational age at delivery for both women with a delivery on the scheduled date and emergency delivery is shown in Figure 2. The mean gestational age at scheduled Cesarean delivery was 38 weeks and 5 days, 57% (121/214) of the women reached this scheduled date. The remaining 43% (93/214) delivered before the scheduled date of Cesarean at a mean gestational age of 35 weeks and 3 days. Of these emergency deliveries 46% (43/93) was performed before 37 weeks of gestation and 22% (20/93) before 32 weeks of gestation. Only two women delivered between 32 and 34 weeks of gestation.

Chapter 5

72

Figure 1. Flowchart of inclusion and exclusion criteria

Figure 3 shows the indications for emergency delivery and gestational age at delivery, 55% (51/93) was due to hemorrhage alone, 36% (33/93) because of early onset of labor and 10% (9/93) because of non-reassuring fetal status.

Table 1 shows the baseline characteristics and the results of the univariable and multivariable analyses. In the univariable analysis, variables associated with emergency delivery were history of Cesarean section, history of premature delivery, maternal age, parity, type of placenta previa, presence of antepartum bleeding, hospital admission because of bleeding, antepartum use of ferrofumarate, and blood transfusion to compensate bleeding.

Based on the multivariable analysis independent predictors for emergency delivery were history of Cesarean section (OR 4.7; 95% CI 1.2-12), presence of antepartum bleeding with one episode (OR 7.5; 95% CI 2.5-23), two episodes (OR 14; 95% CI 4.3-47), three or more episodes (OR 27; 95% CI 8.3-90) and antepartum blood transfusion to compensate bleeding (OR 6.4; 95% CI 1.7-23). Consequently, these variables were used in the final model for the


73

CHAPTER

5

De

liver

y Re

sults

of a

naly

sis

O

n sc

hedu

led

date

Em

erge

ncy

Uni

varia

ble

Mul

tivar

iabl

eTo

tal

n =

121

(56,

5%)

n =

93 (4

3,5%

)O

R (9

5% C

I)p-

valu

eO

R (9

5% C

I)p-

valu

e

Dem

ogra

phic

cha

ract

eristi

csM

ater

nal a

ge, m

ean

(SD)

33.8

(4.5

)32

.6 (5

.3)

1.1

(0.9

9-1.

1)0.

07

Et

hnic

ity, c

auca

sian

(%)

83

(70)

68 (7

4.7)

1.3

(0.7

-2.3

)0.

48

Sm

okin

g, n

(%)

13 (1

1.4)

10 (1

2.2)

1.1

(0.5

-2.7

)0.

87

BM

I, m

ean

(SD)

23.9

(4.8

)24

.4 (6

.4)

1.0

(0.9

6-1.

1)0.

57

Preg

nanc

y ch

arac

teris

tics

Prim

igra

vida

, n (%

)66

(54.

5)61

(65.

6)1.

6 (0

.9-2

.8)

0.1

N

SH

b at

beg

inni

ng o

f pre

gnan

cy, m

ean

in m

mol

/L (S

D)7.

7 (0

.6)

7.6

(0.8

)1.

2 (0

.9-1

.8)

0.27

Ges

tatio

nal a

ge a

t del

iver

y, in

mea

n w

eeks

(IQ

R)38

.6 (3

8.1 -

39.

1)

35.4

(34 -

37.

6)

Ty

pe o

f pla

cent

a pr

evia

to

talis

62 (5

1.2)

56 (6

0.2)

0.89

(0.3

5-2.

2)0.

78

pa

rtial

is16

(13.

2)11

(11.

8)0.

67 (0

.37-

1.2)

0.2

N

Sm

argi

nalis

43 (3

5.5)

26 (2

8)re

fere

nce

Hist

ory

of

Cu

retta

ge29

(24.

0)28

(30.

4)1.

4 (0

.76-

2.6)

0.29

Cesa

rean

sec

tion

13 (1

0.7)

27 (2

9.0)

3.4

(1.6

4-7.

1)0.

001

4.6

(1.8

-11)

<0.0

1U

terin

e su

rger

y5

(4.1

)3

(3.2

)0.

77 (0

.18-

3.3)

0.73

Fibr

oid

tum

ors

6 (5

.0)

3 (3

.2)

0.64

(0.1

6-2.

6)0.

53

Pr

emat

ure

deliv

ery

<37

wee

ks6

(5.0

)14

(15.

1)3.

4 (1

.3-9

.2)

0.01

6

NS

Ante

part

um b

leed

ing

< 3

2 w

eeks

ges

tatio

nal a

ge v

s no

ble

edin

g33

(27.

3)56

(60.

2)18

(6.5

- 50

)<0

.01

*

>32

wee

ks g

esta

tiona

l age

vs

no b

leed

ing

35 (2

8.9)

32 (3

4.4)

9.7

(3.4

-27)

<0.0

1

*Ep

isode

s of

ant

epar

tum

ble

edin

g

Bl

eedi

ng v

s no

ble

edin

g67

(55.

4)88

(94.

6)14

(5.4

-37)

<0.0

0

1

vs n

o bl

eedi

ng38

(31.

4)27

(29.

3)7.

7 (2

.7 -

22)

<0.0

17,

5 (2

.5-2

3)<0

.01

2 vs

no

blee

ding

16 (1

3.2)

23 (2

5.0)

16 (5

.1 -

47)

<0.0

114

(4.3

-47)

<0.0

1>3

vs

no b

leed

ing

13 (1

0.7)

37 (4

0.2)

31 (1

0-94

)<0

.01

27 (8

.3-9

0)<0

.01

Adm

issio

n (d

ays)

bec

ause

of b

leed

ing

1-7

days

vs

no a

dmiss

ion

beca

use

of b

leed

ing

43 (3

5.5)

26 (2

8.3)

6.5

(2.3

-18)

<0.0

1

*>7

day

s v

s no

adm

issio

n be

caus

e of

ble

edin

g24

(19.

8)61

(66.

3)27

(9.8

-77)

<0.0

1

*Co

mpe

nsati

on o

f ant

epar

tum

ble

edin

g

Fe

rrof

umar

ate

37 (3

0.6)

47 (5

1.1)

2.4

(1.4

-4.2

)0.

003

N

SBl

ood

tran

sfus

ion

3 (2

.5)

22 (2

3.9)

12 (3

.6-4

3)<0

.00

6.6

(1.8

-24)

<0.0

1

Abbr

evia

tions

: BM

I bod

y m

ass

inde

x; C

I con

fiden

ce in

terv

al; I

QR

inte

rqua

rtile

rang

e; N

S no

t sig

nific

ant;

OR

odds

ratio

; SD

stan

dard

dev

iatio

n*

Not

incl

uded

in th

e m

ultiv

aria

ble

anal

ysis

beca

use

of c

orre

latio

n

Tabl

e 1.

Bas

elin

e ch

arac

teris

tics

Chapter 5

74

Figure 2. Survival curve of scheduled versus emergency delivery

Before scheduled 93 92 88 86 72 69 60 20 1 0On scheduled 122 122 122 122 122 122 122 99 9 0

prediction of emergency delivery. Comparison of the predictive factors regarding emergency delivery between women with placenta previa who delivered early preterm and late preterm showed no significant differences as shown in table 2.

The area under the ROC curve of the model was 0.832 (95% CI 0.78-0.89). The model showed good agreement between predicted and observed probability. (Figure 4) When divided in five separate risk categories, women in the lowest risk category (i.e. those without an episode of bleeding) almost never needed an emergency delivery, whereas women in the highest risk group virtually all underwent an emergency delivery, as shown in table 3. After internal validation by bootstrapping, the for optimism corrected AUC was 0.821.


75

CHAPTER

5

Table 2. Results of analysis on early preterm and late preterm deliveries

Early preterm delivery < 34 weeks n (%)

Late preterm delivery

34-37weeks n (%) OR (95% CI) p-valueTotal n =22 (51%) n=21 (49%)

Demographic characteristicsAge mother, mean (SD) 33.2 (6.45) 31.5 (4.23) 0.94 (0.84-1.05) 0.29 Pregnancy characteristicsPrimigravida 10 (45.5) 5 (23.8) 0.38 (0.10-1.4) 0.14History of

Cesarean section 7 (31.8) 5 (23.8) 1.5 (0.39-5.7) 0.56Premature delivery 6 (27.3) 3 (14.3) 2.3 (0.48-11) 0.3

Antepartum bleeding < 32 weeks gestational age vs no bleeding 22 (100) 13 (61.9) >32 weeks gestational age vs no bleeding 0 8 (38.1) Episodes of antepartum bleeding

Bleeding vs no bleeding 0 1 (4.8) 1 vs no bleeding 10 (45.5) 8 (38.1) reference2 vs no bleeding 6 (27.3) 5 (23.8) 1.04 (0.23-4.7) 0.96

>3 vs no bleeding 6 (27.3) 7 (33.0) 1.5 (0.35-6.1) 0.61Admission (days) because of bleeding

1-7 days vs >7days 7 (31.8) 5 (23.8) 0.71 (0.19-2.8) 0.63Compensation of antepartum bleeding

Ferrofumarate 7 (31.8) 13 (61.9) 3.5 (0.99-12) 0.052Blood transfusion 8 (36.4) 8 (38.1) 1.08 (0.31-3.7) 0.91

Type of placenta previa complete 11 (50) 15 (71.4) 3,07 (0.75-13) 0.12

partial 2 (9.1) 2 (9.5) 2.3 (0.23-22) 0.49marginal 9 (40.9) 4 (19) reference

Abbreviations: CI confidence interval; OR odds ratio

COMMENTS

Main findingsIn this study, we investigated predictors for emergency delivery in women with placenta previa. The incidence of placenta previa was 0.44%. In 43% of the women, the delivery was emergent, at a mean gestational age of 35 weeks and 3 days. In multivariable regression, we identified a history of Cesarean section, number of episodes of antepartum bleeding and need for blood transfusion as independent predictors for an emergency delivery. A prediction model including these predictors showed good calibration and discrimination.

Chapter 5

76

Table 3. Predictors by risk category

* Risk for emergency delivery per category, respectively: 8%, 27%, 50%, 62%, 87%

Risk category* N Previous CS (%)

Need for blood-

transfusion (%)

Episodes of bleeding (%)

1 2 >3

1 49 0 0 0 0 02 59 12 (20) 0 49 (83) 0 03 26 0 0 0 26 (100) 04 39 9 (23) 0 9 (23) 0 30 (77)5 39 21 (54) 25 (64) 7 (18) 12 (31) 20 (51)

Figure 3. Survival curve for indications for emergency delivery

Hemorrhage 51 51 49 49 41 40 39 13 0 0Labor 33 33 31 30 26 24 20 6 1 0NRFS 9 8 8 7 5 5 1 1 0 0


77

CHAPTER

5

Figure 4. Calibration plot

Strengths and limitationsOur study is a multicenter cohort study reducing selection bias. Another strength of our study is the definition of antepartum bleeding to forestall variations in management; several different variables (hemorrhage leading to admission, the duration of admission, number of episodes of antepartum bleeding, and interventions needed to compensate) were extracted from the medical charts to objectify the amount of antepartum bleeding.

The timing of elective Cesarean delivery as well as the decision to perform an emergency Cesarean delivery are interventions of which the final decision is made by attending physicians. Because of the retrospective character and the absence of a prospective protocol, our study could not completely control for the varying management in women with symptomatic placenta previa. Although we tried to forestall this as much as possible by performing a multicenter study, the variability and the different strategies are present and a limitation in our study.

Chapter 5

78

Furthermore, we used the results of the most recent ultrasound to confirm the diagnosis of placenta previa, since a second trimester placenta previa resolves in approximately 90% during pregnancy.12 This comes with a limited predictive value. On the contrary, the 32 weeks scan –for evaluation of placental localization in women with placenta previa– as confirmation of the diagnosis will not be available for all women, since around 9% of them already delivered by that time. Regarding our model for the prediction of emergency delivery in women with placenta previa, internal validation was assessed by bootstrapping and the model retained discriminatory capacity and goodness of fit in the validation samples. However, we do not possess of an external dataset for external validation.

InterpretationThe incidence of placenta previa found in our study is similar to the incidences reported in previous studies.2, 13 From previous studies, it is already known that antepartum bleeding is a strong predictor for emergency delivery in women with placenta previa. However, in these retrospective cohort studies by Fishman et al.5 and by Lam et al.6 criteria for antepartum bleeding were not or insufficiently defined, such as currently in our study. A recent study by Luangruangrong et al14 also found an increased risk of preterm delivery in women with antepartum bleeding.

We found a history of Cesarean section as predictor for emergency delivery. A previous Cesarean section is a well-known risk factor for placenta previa, but to our knowledge, it has not been described as a predictor for emergency delivery before, and future studies are needed to confirm this finding. We hypothesize that a placenta overlying a Cesarean scar might be prone to antepartum bleeding because of a different placental adhesion to scar tissue than to normally developed endometrium. Potentially, the accrete placenta is a risk factor for antepartum bleeding rather than the history of Cesarean section itself. Although accrete placenta was equally distributed in our study cohort, the total count of this complication was low and our study is underpowered to analyze this hypothesis.

We found no difference in the frequency of emergency delivery in complete and incomplete placenta previa as was also concluded by Tuzovic et al.13 but is in contrast with the results of Dola et al.15 Differences in these conclusions may be explained by differences in used diagnostic techniques, a broad variation in study periods and clinical management. Cervical shortening as a predictive factor for preterm delivery in women with symptomatic placenta previa and antepartum bleeding has been reported. 5, 16 It seems plausible that a shortening cervix increases the risk of preterm delivery, but we could not take this variable into account since cervical length was not routinely measured in the Netherlands during the study period.


79

CHAPTER

5

The 2011 NIHCD consensus conference publication on late preterm and early term delivery advises to conduct the delivery between 36 and 37 weeks gestation in women with placenta previa. Our cohort consists of data before this publication with Cesarean deliveries mainly scheduled after 38 weeks of gestation. A recent study by Balayla et al.17 showed early-term delivery in placenta previa to be associated with fewer complications and no greater risk than late preterm delivery. Considering our results, of all women that were scheduled for Cesarean delivery after 37 weeks of gestation, 63% reached this date. This suggests that with adequate selection, iatrogenic premature delivery may potentially be reduced. For example, the majority of women in our study who had emergency delivery before 37 weeks of gestation had antenatal bleeding in the second trimester. Potentially women can be treated conservatively in an outpatient setting in the absence of bleeding before 32 weeks of gestation.

Finally, because three or more episodes of antenatal bleeding indicate a very high risk (87%) of emergency delivery, scheduling delivery before 37 weeks of gestation with fetal lung maturation could be considered in these women to decrease neonatal respiratory distress syndrome.18 Since the effect of antenatal corticosteroids decreases after seven days of administration19 and repeat doses of antenatal corticosteroids are associated with increased rates of behavioural disorders, timing is very important and we should restraint from routine administration of corticosteroids at every episode of antepartum bleeding.20, 21

CONCLUSION

Our study results should be evaluated for application in individualized antenatal care concerning timing of delivery and administration of corticosteroids. This can potentially lead to conservative treatment in an outpatient setting in women at low risk for emergency delivery for the reason of beneficial neonatal outcome by reducing iatrogenic preterm delivery.

Chapter 5

80

REFERENCES

1. Johnston T.A.., S. P-B. Placenta Praevia, Placenta Praevia Accreta and Vasa Praevia: Diagnosis and Management. 2011 [cited; Available from:

2. Ananth CV, Smulian JC, Vintzileos AM. The effect of placenta previa on neonatal mortality: a population-based study in the United States, 1989 through 1997. American journal of obstetrics and gynecology. 2003 May;188(5):1299-304.

3. Zlatnik MG, Cheng YW, Norton ME, Thiet MP, Caughey AB. Placenta previa and the risk of preterm delivery. The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2007 Oct;20(10):719-23.

4. Crane JM, Van den Hof MC, Dodds L, Armson BA, Liston R. Maternal complications with placenta previa. Am J Perinatol. 2000;17(2):101-5.

5. Fishman SG, Chasen ST, Maheshwari B. Risk factors for preterm delivery with placenta previa. J Perinat Med. 2012 Jan;40(1):39-42.

6. Lam CM, Wong SF, Chow KM, Ho LC. Women with placenta praevia and antepartum haemorrhage have a worse outcome than those who do not bleed before delivery. J Obstet Gynaecol. 2000 Jan;20(1):27-31.

7. Onwere C, Gurol-Urganci I, Cromwell DA, Mahmood TA, Templeton A, van der Meulen JH. Maternal morbidity associated with placenta praevia among women who had elective caesarean section. European journal of obstetrics, gynecology, and reproductive biology. 2011 Nov;159(1):62-6.

8. Crane JM, van den Hof MC, Dodds L, Armson BA, Liston R. Neonatal outcomes with placenta previa. Obstetrics and gynecology. 1999 Apr;93(4):541-4.

9. Norgaard LN, Pinborg A, Lidegaard O, Bergholt T. A Danish national cohort study on neonatal outcome in singleton pregnancies with placenta previa. Acta obstetricia et gynecologica Scandinavica. 2012 May;91(5):546-51.

10. (NVOG) DAoOaG. Guideline Blood loss in the second half of pregnancy. 2011 [cited; Available from:

11. Smith GC, Seaman SR, Wood AM, Royston P, White IR. Correcting for optimistic prediction in small data sets. Am J Epidemiol. 2014 Aug 1;180(3):318-24.

12. Neilson JP. Interventions for suspected placenta praevia. The Cochrane database of systematic reviews. 2003(2):CD001998.

13. Tuzovic L. Complete versus incomplete placenta previa and obstetric outcome. Int J Gynaecol Obstet. 2006 May;93(2):110-7.

14. Luangruangrong P, Sudjai D, Wiriyasirivaj B, Paloprakarn C. Pregnancy outcomes of placenta previa with or without antepartum hemorrhage. J Med Assoc Thai. 2013 Nov;96(11):1401-7.

15. Dola CP, Garite TJ, Dowling DD, Friend D, Ahdoot D, Asrat T. Placenta previa: does its type affect pregnancy outcome? Am J Perinatol. 2003 Oct;20(7):353-60.

16. Ghi T, Youssef A. Ultrasonographic cervical length and risk of hemorrhage in pregnancies with placenta previa. Obstetrics and gynecology. 2010 Dec;116(6):1458; author reply -9.

17. Balayla J, Wo BL, Bedard MJ. A late-preterm, early-term stratified analysis of neonatal outcomes by gestational age in placenta previa: defining the optimal timing for delivery. The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2015;28(15):1756-61.

18. Roberts D, Dalziel S. Antenatal corticosteroids for accelerating fetal lung maturation for women at risk of preterm birth. The Cochrane database of systematic reviews. 2006(3):CD004454.


81

CHAPTER

5

19. Melamed N, Shah J, Soraisham A, Yoon EW, Lee SK, Shah PS, et al. Association Between Antenatal Corticosteroid Administration-to-Birth Interval and Outcomes of Preterm Neonates. Obstetrics and gynecology. 2015 Jun;125(6):1377-84.

20. French NP, Hagan R, Evans SF, Mullan A, Newnham JP. Repeated antenatal corticosteroids: effects on cerebral palsy and childhood behavior. American journal of obstetrics and gynecology. 2004 Mar;190(3):588-95.

21. Stutchfield PR, Whitaker R, Gliddon AE, Hobson L, Kotecha S, Doull IJ. Behavioural, educational and respiratory outcomes of antenatal betamethasone for term caesarean section (ASTECS trial). Arch Dis Child Fetal Neonatal Ed. 2013 May;98(3):F195-200.

CHAPTER 6

Incidence and recurrence rate of placental abruption:

a longitudinal linked national cohort study in the

Netherlands

Ruiter L Ravelli ACJ

De Graaf IMMol BWJ

Pajkrt E

American Journal of Obstetrics and Gynecology. 2015 Oct;213(4):573.e1-8

Chapter 6

84

ABSTRACT

Objective Women who have experienced a placental abruption have a risk of recurrence, but exact information to quantify this risk is currently not available. We studied the incidence and recurrence rate of placental abruption in a subsequent pregnancy and the influence of hypertensive disorders.

Study design We conducted a retrospective national cohort study of all singleton pregnancies that ended between 1999 and 2007 in the Netherlands. A longitudinal linked national cohort of these women with information on a subsequent singleton delivery was used. We calculated and compared incidence and recurrence rates of placental abruption for women in total, stratified by gestational age of first placental abruption and by the presence of a hypertensive disorder in their first pregnancy.

Results We studied 1,570,635 women of which 3,496 (0.22%) experienced a placental abruption. Information was available on a subsequent singleton delivery for 264,424 deliveries. Of these, 521 women (0.20%) had a placental abruption in the first pregnancy versus 214 (0.08%) in the second pregnancy. The risk of placental abruption in a subsequent pregnancy was significantly higher in women with a previous placental abruption compared with women without (5.8% vs 0.06%; adjusted odds ratio (aOR) 93; 95% confidence interval (CI) 62-139). Women with a placental abruption that occurred at term in their first pregnancy were more at risk for recurrence (aOR 188; 95% CI 116-306) than women with a preterm (aOR 52; 95% CI 25-111) or early preterm (< 32 weeks) placental abruption in their first pregnancy (aOR 39; 95% CI 13-116). Placental abruption was more frequent among women with a hypertensive disorder compared with normotensive women (0.44% versus 0.16%, OR 2.7; 95%CI 2.3 – 3.3). Women with a hypertensive disorder were less at risk for recurrence than were normotensive women (aOR 0.68; 95% CI 0.27-1.6). No interaction between a hypertensive disorder in the first pregnancy and the recurrence risk was found.

Conclusion Women with a placental abruption in their first pregnancy have a greatly increased risk of placental abruption in a subsequent pregnancy. Hypertensive disorders increase the risk of placental abruption but do not increase the recurrence rate in a subsequent pregnancy. We suggest elective induction from 37 weeks for women with a history of placental abruption at term in a previous pregnancy.

Incidence and recurrence rate of placental abruption

85

CHAPTER

6

INTRODUCTION

Placental abruption, the premature dehiscence of a normally inserted placenta from the uterine wall, is associated with significant perinatal morbidity and accounts for 12% of all perinatal deaths in the United States.1 Early diagnosis of placental abruption potentially can improve the outcome of the children involved in these pregnancies by closely monitoring women at risk. 2 Advanced maternal age, multiparity 3, chronic hypertension 3, smoking 4, 5, cocaine use 6, previous cesarean delivery 5, 7, uterine surgery 5 and short interpregnancy interval 7 are reported risk factors for placental abruption. In patients who experienced a placental abruption in a previous pregnancy, there seems to be a tendency for recurrence, although most studies on the incidence and recurrence rate of placental abruption are dated or include relatively small samples of patients.8, 9 Knowledge of these subjects potentially can help in providing specialized prenatal care for women who are at risk for placental abruption for the reason of beneficial outcome.10 Our objective was to investigate the incidence and recurrence rate of placental abruption in a retrospective study of two subsequent deliveries in the Netherlands between 1999 and 2007. Moreover, we assessed the influence of hypertensive disorders on the incidence and recurrence rate.

MATERIALS AND METHODS

This study was performed in a retrospective nationwide cohort with the use of the Netherlands Perinatal Registry (PRN). The PRN consists of prospectively collected population-based data that contain information on pregnancies, deliveries and (re)admissions until 28 days after birth. The database is obtained by a validated linkage of three different registries: the midwifery registry (LVR1), the obstetrics registry (LVR2), and the neonatology registry (LNR) of hospital admissions of newborn infants.11, 12 The coverage of the PRN registry is approximately 96% of all deliveries in the Netherlands and it contains all pregnancies that ended in live births and stillbirths at > 22 weeks of gestation and a birthweight of at least 500 g. Furthermore, all admissions to the neonatology care unit are registered until 28 days after birth. It is used primarily for annual assessment of the quality indicators of obstetric care. The records included in het PRN registry are entered at child’s level. There is no unique maternal identifier available in the registry to follow-up on outcomes of subsequent pregnancies of the same mother. Previously, a longitudinal probabilistic-linkage procedure in which records of children of the same mother are linked was performed to create a mother identifier. This resulted in a cohort of 272,551 women with complete data on first and second

Chapter 6

86

deliveries of the same mother. Details on this linkage procedure are described elsewhere.13 The data in the PRN database are anonymous, the Dutch Perinatal Registry gave their approval for the use of the data for this study (approval no. 12.71).

Inclusion and exclusion criteriaWe selected all women who delivered in The Netherlands between Jan. 1, 1999 and Dec. 31, 2007. We excluded women with a multiple gestation (n = 7,472) and women with a gestational age at delivery of > 42 + 6 weeks (n = 655).

Outcome measuresOur primary outcome was placental abruption. We identified all women with a placental abruption that had been registered in the PRN registry with a code for placental abruption. Codes are assigned by midwives, residents and obstetricians based on information in the patients’ medical record. A clinical diagnosis of placental abruption is made based on a combination of the following clinical symptoms: vaginal bleeding, abdominal pain or fetal distress and further supported by ultrasonographic findings as retro placental bleeding. After delivery, the diagnosis may be confirmed by a couvelaire uterus, gory amniotic fluid, or examination of the placenta that reveals a clot.14 Placental abruption is primarily documented by one of the care providers after delivery; there is no stratification in severity in the PRN database. In addition, all women with a positive history of placental abruption at the start of the second pregnancy were labelled as cases of placental abruption in the first pregnancy. The gestational age at delivery of the fetus was used as the gestational age of placental abruption.

CharacteristicsWe extracted demographic and obstetric baseline characteristics from the PRN registry as ethnicity, maternal age, socioeconomic status, pregnancy interval, hypertension, and neonatal sex. All characteristics were primarily documented by the care providers. Ethnicity was categorized into white (native Dutch and other Caucasian women) and non-white (different ethnic groups as African/Surinamese, South Asain, Moroccan and Turkish) for this study. The continuous socio economic status score (based on mean income level, the percentage of households with a low income, inhabitants without a paid job, and households with, on average, a low education in a postal code area) was categorized into a high, middle, and low group based on percentile ranges (≤ 25th percentile, middle, > 75th percentile). Gestational age was based on last menstrual period or ultrasound measurements at < 20 weeks. If the estimation by ultrasound measurement differed > 6 days from the last menstrual period, then the ultrasound measurement was considered the dominant measurement. Hypertensive disorders (HTD) included chronic hypertension, pregnancy-induced hypertension, and


87

CHAPTER

6

preeclampsia. The clinical diagnosis of hypertension was made with a systolic blood pressure of ≥ 140 mm Hg and/or a diastolic blood pressure of ≥ 90 mm Hg. Preeclampsia was diagnosed with hypertension and proteinuria.15 Cases of placental abruption were analyzed in total and stratified into three groups based on the gestational age at the time of placental abruption in the first pregnancy; early preterm (22+0- 31+6 weeks of gestation), late preterm (32+0- 36+6 weeks of gestation) and term (37+0- 42+6 weeks of gestation). In addition, we stratified the analysis by HTD in the first pregnancy: women with an HTD in their first pregnancy and women without an HTD in their first pregnancy.

AnalysisWe calculated the incidence of placental abruption in our cohort of 1,570,635 singleton deliveries by dividing the number of placental abruptions by the total number of deliveries. We calculated the incidence per gestational week by dividing the number of placental abruptions in a particular week by the total number of deliveries in that given week. We selected all women with a subsequent singleton delivery and compared incidence of placental abruption in a subsequent pregnancy in women with and without a placental abruption in their first pregnancy. In addition, we compared their demographic and obstetric baseline characteristics. Comparison of the baseline characteristics was by univariate analysis that was performed with the unpaired t-test for normally distributed continuous variables and with the χ2 test for categoric variables. All tests were two-sided with a probability value of .05 as threshold for statistical significance. To estimate the effect of a history of placental abruption on the occurrence of placental abruption in a subsequent pregnancy, logistic regression modeling was used and was expressed as odds ratio (OR) with a 95% confidence interval (CI). In addition, the effect of gestational age at placental abruption in the first pregnancy was estimated and expressed as odds ratio with 95% CI. To account for the effect of potential confounders, we used a multivariate logistic regression analysis to adjust for variables that were unequally distributed in the baseline characteristics. When a low incidence of placental abruption was found, Firths’ penalized likelihood approach was performed to minimize small sample size bias. Kaplan-Meier curves by gestational age of first placental abruption were constructed for the recurrent placental abruptions. We tested for an association between placental abruption and HTD in the first pregnancy. When statistically significant, analyses were also performed separately for HTD and non-HTD cases per three strata of gestational age of first placental abruption. In addition, we tested for interaction between HTD in the first pregnancy and the recurrence of placental abruption in a subsequent pregnancy by adding an interaction term to our logistic regression model. The data were analyzed with the SAS statistical software package (version 9.3; SAS Institute Inc, Cary, NC, USA).

Chapter 6

88

RESULTS

Between Jan. 1, 1999 and Dec. 31, 2007, a total of 1,570,635 singleton pregnancies were identified, 3,496 pregnancies (0.22%) were complicated by a placental abruption. The overall incidence was highest between 28 and 31 weeks of gestation; 40.8 per 1000 deliveries (4.1%). Between 32 and 34 weeks of gestation, the incidence was 20.5 per 1000 deliveries (2.1%); between 39 and 42 weeks of gestation the incidence was 0.8 per 1000 (0.08%). (Figure 1).

Figure 1. Incidence of placental abruption among singletons per 1000 deliveries by gestational age

This figure shows the incidence of placental abruption per 1000 deliveries per gestational week of pregnancy.

A total of 264,424 women (528,848 deliveries) were available for the analysis of the recurrence rate and incidence of placental abruption in a subsequent pregnancy. Baseline characteristics of our linked cohort are listed in Table 1. In the first pregnancy, 521 women (0.20%) had a placental abruption, among these women an age < 25 years, non-white ethnicity, low socioeconomic status and HTDs were more prevalent. Of the 521 women with a placental abruption in the first pregnancy, 30 women (5.8%) had a recurrent placental abruption in the subsequent pregnancy. Placental abruption in a subsequent pregnancy occurred in 184 parous women (0.07%) without a history of placental abruption. Figure 2 shows all women with a placental abruption in the first pregnancy and their time to a recurrent placental abruption stratified by gestational age of the first placental abruption.


89

CHAPTER

6

Table 1. Baseline characteristics of the linked cohort

Characteristics of the first pregnancy Placental abruption in first pregnancy

(n = 521)

No placental abruption in first pregnancy

(n = 263,903)

p-value

Maternal characteristics n % n %

Mean maternal age, yearsA 28.0 ± 4.6 28.6 ± 4.2 0.0001Maternal age < 25 years 115 22,1 42.784 16,2 0.0003Mean gestational age, weeksA 34.6 ± 4.3 39.2 ± 2.2 < 0.0001Non-white maternal ethnicity (n) 74 14,2 28,826 10,9 0.0165Low socio-economic status (n) 149 28,6 57,732 21,9 0.0002Hypertensive disorder 147 28,2 33,186 12,6 <0.0001Fetal characteristics, male (n) 283 54,3 135,887 51,5 0.1970

A Presented as mean ± SD

Figure 2. Recurrent placental abruption by gestational age of first placental abruption

This figure shows the time to placental abruption in a subsequent pregnancy after an first early preterm, late preterm or term placental abruption.

After adjustment for maternal age, ethnicity, socioeconomic status, and hypertension, a placental abruption in the first pregnancy was associated with a significantly increased risk of placental abruption in a subsequent pregnancy (adjusted odds ratio (aOR) 93, 95%CI 62-139). Table 2 shows the distribution of placental abruption in a subsequent pregnancy in women with and without a history of placental abruption and the crude and adjusted odds ratios of placental abruption in the first pregnancy on the recurrence rate of placental abruption in

Chapter 6

90

a subsequent pregnancy, stratified by gestational age of first placental abruption. The odds ratio of a history of placental abruption on the recurrence in a subsequent pregnancy was higher for women with a placental abruption in the term period of the first pregnancy than for women with a previous late or early preterm placental abruption.

Table 2. Placental abruption in a subsequent pregnancy in total and by gestational age of placental abruption in the first pregnancy

Placental abruption in second pregnancy Total abruption

second pregnancy

Odds ratio and 95% confidence interval

22+0 - 31+6 32+0 - 36+6 37+0 - 42+6

n n n n % Crude AdjustedA

Complete cohort of women with first term delivery, no placental abruption in first pregnancy (n=243,799)

26 62 70 158 0,1% = Reference =

Complete cohort placental abruption first pregnancy (n = 521)

2 13 15 30 5,8% 95 (64-142)

93 (62-139)

Gestational age of placental abruption in the first pregnancy:

22+0 - 31+6 weeks (n=128) 1 1 1 3 2,3% 43 (16-126) 39 (13-116)

32+0 - 36+6 weeks (n=210) 0 4 3 7 3,3% 57 (27-120) 52 (25-111)

37+0 - 42+6 weeks (n=183) 1 8 11 20 11% 193 (119-313)

188 (116-306)

AAdjusted for hypertension, maternal age, ethnicity, low SES

Of the 264,424 first singleton pregnancies, a total of 33,333 women (12.6%) had an HTD. In the second pregnancy, 17,378 (6.6%) women had an HTD. In the first pregnancy, placental abruption was more prevalent in the HTD group compared with the non-HTD group (0.44% vs 0.16%; OR 2.73, 95%CI 2.26-3.31). In the second pregnancy, placental abruption was also more prevalent in the HTD group compared to the non-HTD group (0.52% vs 0.17%). Table 3 shows that, in both women with and without HTD in their first pregnancy, the recurrence risk of placental abruption is increased for women with a history of placental abruption compared to women without a history of placental abruption. Table 3 also shows that the odds ratio of a history of placental abruption on the recurrence risk for is lower for women with an HTD in their first pregnancy compared with women who were normotensive in their first pregnancy. Although the initial risk of placental abruption is increased by the


91

CHAPTER

6

presence of an HTD, the interaction between an HTD in the first pregnancy and the recurrence risk of placental abruption in a subsequent pregnancy was not statistically significant (p = 0.1).

Table 3. Placental abruption in a subsequent pregnancy by presence of HTD in the first pregnancy

Complete cohort placental abruption first pregnancy (n = 521)

Placental abruption in second pregnancy Total abruption

second pregnancy

Odds ratio and 95% confidence interval

22+0 - 31+6 32+0 - 36+6 37+0 - 42+6

n n n n % Crude AdjustedB

HTD in first pregnancy (n = 147) 2 2 4 6 4,1% 0.66

(0.27 - 1.60)0.68

(0.28 - 1.63)

No HTD in first pregnancy (n = 347) 0 11 11 24 6,9% = Reference =

B Adjusted for maternal age, ethnicity, low SES

COMMENT

In this study, we investigated the overall incidence and recurrence rate of placental abruption among singleton pregnancies. Moreover, we investigated the influence of HTD on placental abruption. The incidence of placental abruption among singleton pregnancies was 0.22%. The incidence was higher in women with an HTD compared with normotensive women. The risk of placental abruption in a subsequent pregnancy was significantly higher in women with a previous placental abruption, compared with women without a previous placental abruption. Women with a first placental abruption at term were most at risk for recurrence. Our study has both strengths and limitations. Because placental abruption is a rare, but important, event, large amounts of data are needed to detect differences in populations. Our study was performed with data from a large population-based national registry in which data are collected prospectively. This provides validity of the results because it represents a national outcome, which avoids recall and reporting bias by prospective collection and avoids selection bias that is created potentially by hospital-based studies. However, this study was a retrospective study, which comes with limitations. We needed to rely on the available information in the PRN; thus, we cannot rule out misclassification bias. It is likely that some cases have been missed, especially the mild or concealed ones. Another limitation is the unavailability of gestational age of abruption. From 32 weeks of gestation, placental abruption will mostly be followed by immediate delivery. Placental abruption can

Chapter 6

92

also be managed expectantly when occurring at <32 weeks of gestation. The gestational age of abruption is then different than the date of delivery. The PRN registry does not provide this information; therefore, the difference in preterm and term incidence of placental abruption may be higher than found in this study. Data on possible confounders as body mass index and tobacco use are also not available in the PRN, and the usage of a linked database comes with restrictions as no linkage because of missing values of the linkage variables. The incidence of placental abruption described in previous studies is 0.3-1%, which is higher than in our study.3-5 There are several potential explanations for this difference, for example missing cases due to the retrospective character of this study as mentioned before, or inconsistencies in diagnostic criteria of placental abruption; there are neither laboratory tests nor diagnostic methods to accurately detect placental abruption. A previous study revealed that inconsistencies in the composition of placental abruption cases lead to differences in incidence.3 Most studies on incidence include singleton and multiple pregnancies. The latter are associated with a higher incidence of placental abruption, thus exclusion of multiple pregnancies from our analysis may have led to a lower incidence.16 With the exclusion of subsequent multiple pregnancies we potentially excluded some women with a history of placental abruption as well. Furthermore, the incidence of placental abruption per 1000 deliveries decreases by gestational age. Some studies included pregnancies with a minimum gestational age of 16 weeks, whereas the PRN-registry comprises only pregnancies from a gestational age of 22 weeks. 17-19

In general, the incidence of placental abruption seems to be lower in Nordic countries than in the United States, where the population consists of more women from African American ethnicity. These women are more at risk for placental abruption than white women because HTD, smoking, alcohol and cocaine use are more prevalent in this population.20-23 Our study population is relatively more homogenous consisting of 90% maternal white ethnicity, thus potentially leading to a lower incidence. Furthermore, the percentage of Cesarean section and thus the number of women with a history of Cesarean section is significantly lower in our country compared with the United States.24

Studies in the United States found an increasing incidence of placental abruption over time, in contrast to Tikkanen et al,25 who performed a large population-based cohort study in Finland and found a decrease in the incidence of placental abruption from 1980 – 2005. Because we calculated the incidence on a recent cohort in the Netherlands including mainly deliveries in the 21st century, this partially might explain our slightly lower incidence than studies performed from 1980-2000. Also, improvements in prenatal care over the last decades may have contributed. Women with HTD, preeclampsia, or small-for-gestational-age fetuses are diagnosed earlier and delivered before their due date, which potentially could decrease


93

CHAPTER

6

the risk of a placental abruption. Unfortunately, studies, for example, on medical decision-making in pregnancy-induced hypertension or preeclampsia frequently are underpowered to find significant effects on placental abruption.25, 26 On the other hand, several risk factors for placental abruption, such as advanced maternal age and a previous cesarean section, are on the rise; therefore the exact development of incidence of placental abruption remains difficult to determine. The decrease in incidence of placental abruption with advancing gestation in our study has been found in previous studies.17-19, 25, 27 Our distribution of incidence by gestational age is similar to Tikkanen et al.25

The overall recurrence rate of placental abruption in this study was 5.8% (range 2.3-11%). An increased risk for placental abruption after a previous placental abruption is found in previous studies, although in less well defined populations, and data of these studies are dated.3, 9, 28-35 The true recurrence rate may even be higher than found in our study because women who have had placental abruption are less likely than women without a history of placental abruption to have another pregnancy.8 In addition, elective delivery at term before labor in women with a history of placental abruption may also distort the true recurrence rate because, in these women, it is not known whether a placental abruption would have occurred. On the contrary, obstetricians potentially are more likely to diagnose and correctly code a placental abruption in women with a history of placental abruption compared with women without this history, which can amplify the recurrence rate. To the best of our knowledge, only one previous study investigated the influence of gestational age at first placental abruption on the recurrence rate.2 Rasmussen et al suggested that women with a history of placental abruption should receive special surveillance six weeks before the gestational age of the first placental abruption to reduce the recurrence risk to the initial risk. This is in line with our results that showed that almost 60% of the second placental abruptions occurred in the same gestational period and that 40% of all women with placental abruption in the term period of the first pregnancy experienced recurrence in the late preterm period. We may add to these results that primarily women with first-term placental abruption are at risk for recurrence (11%). Thus, even though the incidence of a placental abruption in the preterm period is higher than the incidence of placental abruption at term, the consequences for a subsequent pregnancy seem to be lower. However, we must interpret these results with caution because it is likely that term placental abruptions are diagnosed and reported more accurately than early preterm abruptions. The association between HTD and placental abruption has been found in previous studies.3, 36 The absent interaction between HTD and the recurrence of placental abruption in a subsequent pregnancy that was found in our study has not been described before. From our results, it seems that placental abruptions in normotensive women are more likely to recur than placental abruptions in women with an HTD. This possibly may be explained partially by the fact that the prevalence of HTD is less in second than in first pregnancies, attenuating

Chapter 6

94

this as a risk factor.37, 38 It might also be due to more intensive monitoring of women in subsequent pregnancies after a previous pregnancy-induced hypertension or preeclampsia, leading to earlier diagnosis of complications and delivery. It would be valuable to stratify recurrence risk of placental abruption by type of HTD, because chronic hypertension and severe pre-eclampsia are stronger predictors for placental abruption than pregnancy-induced hypertension and mild pre-eclampsia; however, the design of the PRN registry did not allow us to perform such analysis.3, 39 Furthermore, we would suggest elective induction at 37 weeks of gestation in women with a previous placental abruption in the term period of pregnancy. Although the design of this study cannot provide a clear answer to the question on the best strategy in these women and multiple factors influence delivery timing beyond just the recurrence probability of placental abruption, more than 50% of the recurrent abruptions occurred after 37 weeks of gestation. In conclusion, the findings of this study show that women with placental abruption are at increased risk for recurrence, especially when the first placental abruption occurred in the term period. HTDs increase the initial risk of placental abruption. Potentially, normotensive women with a placental abruption in their first pregnancy are more at risk for recurrence than are women with an HTD in their first pregnancy. It is desirable to monitor pregnant women with a history of placental abruption in tertiary centers and to closely monitor them in the gestational period of first placental abruption.

ACKNOWLEDGEMENTS

We thank all Dutch midwives, obstetricians, neonatologists and other perinatal health care providers for the registration of perinatal information and the Foundation of The Netherlands Perinatal Registry (www.perinatreg.nl) for permission to use the registry data.


95

CHAPTER

6

REFERENCES

1. Ananth CV, Wilcox AJ. Placental abruption and perinatal mortality in the United States. American journal of epidemiology. 2001 Feb 15;153(4):332-7.

2. Rasmussen S, Irgens LM, Albrechtsen S, Dalaker K. Women with a history of placental abruption: when in a subsequent pregnancy should special surveillance for a recurrent placental abruption be initiated? Acta Obstet Gynecol Scand. 2001 Aug;80(8):708-12.

3. Ananth CV, Savitz DA, Williams MA. Placental abruption and its association with hypertension and prolonged rupture of membranes: a methodologic review and meta-analysis. Obstetrics and gynecology. 1996 Aug;88(2):309-18.

4. Ananth CV, Savitz DA, Luther ER. Maternal cigarette smoking as a risk factor for placental abruption, placenta previa, and uterine bleeding in pregnancy. American journal of epidemiology. 1996 Nov 1;144(9):881-9.

5. Tikkanen M, Nuutila M, Hiilesmaa V, Paavonen J, Ylikorkala O. Prepregnancy risk factors for placental abruption. Acta Obstet Gynecol Scand. 2006;85(1):40-4.

6. Hulse GK, Milne E, English DR, Holman CD. Assessing the relationship between maternal cocaine use and abruptio placentae. Addiction. 1997 Nov;92(11):1547-51.

7. Getahun D, Oyelese Y, Salihu HM, Ananth CV. Previous cesarean delivery and risks of placenta previa and placental abruption. Obstetrics and gynecology. 2006 Apr;107(4):771-8.

8. Rasmussen S, Irgens LM, Dalaker K. The effect on the likelihood of further pregnancy of placental abruption and the rate of its recurrence. Br J Obstet Gynaecol. 1997 Nov;104(11):1292-5.

9. Karegard M, Gennser G. Incidence and recurrence rate of abruptio placentae in Sweden. Obstetrics and gynecology. 1986 Apr;67(4):523-8.

10. Oyelese Y, Ananth CV. Placental abruption. Obstetrics and gynecology. 2006 Oct;108(4):1005-16.11. Meray N, Reitsma JB, Ravelli AC, Bonsel GJ. Probabilistic record linkage is a valid and transparent

tool to combine databases without a patient identification number. Journal of clinical epidemiology. 2007 Sep;60(9):883-91.

12. Tromp M, Ravelli AC, Meray N, Reitsma JB, Bonsel GJ. An efficient validation method of probabilistic record linkage including readmissions and twins. Methods of information in medicine. 2008;47(4):356-63.

13. Schaaf JM, Hof MH, Mol BW, Abu-Hanna A, Ravelli AC. Recurrence risk of preterm birth in subsequent singleton pregnancy after preterm twin delivery. American journal of obstetrics and gynecology. 2012 Oct;207(4):279 e1-7.

14. Guideline Blood loss in the second and third trimester of pregnancy. Utrecht: The Dutch Society of Obstetrics and Gynecology; 2008.

15. De Paepe ME, Shapiro S, Hanley LC, Chu S, Luks FI. Correlation between cord insertion type and superficial choriovasculature in diamniotic-monochorionic twin placentas. Placenta. 2011 11/2011;32(11):901-5.

16. Ananth CV, Smulian JC, Demissie K, Vintzileos AM, Knuppel RA. Placental abruption among singleton and twin births in the United States: risk factor profiles. Am J Epidemiol. 2001 Apr 15;153(8):771-8.

17. Sheiner E, Shoham-Vardi I, Hadar A, Hallak M, Hackmon R, Mazor M. Incidence, obstetric risk factors and pregnancy outcome of preterm placental abruption: a retrospective analysis. The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2002 Jan;11(1):34-9.

Chapter 6

96

18. Sheiner E, Shoham-Vardi I, Hallak M, Hadar A, Gortzak-Uzan L, Katz M, et al. Placental abruption in term pregnancies: clinical significance and obstetric risk factors. The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2003 Jan;13(1):45-9.

19. Rasmussen S, Irgens LM, Bergsjo P, Dalaker K. The occurrence of placental abruption in Norway 1967-1991. Acta obstetricia et gynecologica Scandinavica. 1996 Mar;75(3):222-8.

20. Ananth CV, Berkowitz GS, Savitz DA, Lapinski RH. Placental abruption and adverse perinatal outcomes. JAMA. 1999 Nov 3;282(17):1646-51.

21. Ananth CV, Oyelese Y, Yeo L, Pradhan A, Vintzileos AM. Placental abruption in the United States, 1979 through 2001: temporal trends and potential determinants. American journal of obstetrics and gynecology. 2005 Jan;192(1):191-8.

22. Pritchard JA, Cunningham FG, Pritchard SA, Mason RA. On reducing the frequency of severe abruptio placentae. American journal of obstetrics and gynecology. 1991 Nov;165(5 Pt 1):1345-51.

23. Nilsen RM, Vollset SE, Rasmussen SA, Ueland PM, Daltveit AK. Folic acid and multivitamin supplement use and risk of placental abruption: a population-based registry study. Am J Epidemiol. 2008 Apr 1;167(7):867-74.

24. American College of O, Gynecologists, Society for Maternal-Fetal M, Caughey AB, Cahill AG, Guise JM, et al. Safe prevention of the primary cesarean delivery. American journal of obstetrics and gynecology. 2014 Mar;210(3):179-93.

25. Tikkanen M, Riihimaki O, Gissler M, Luukkaala T, Metsaranta M, Andersson S, et al. Decreasing incidence of placental abruption in Finland during 1980-2005. Acta obstetricia et gynecologica Scandinavica. 2012 Sep;91(9):1046-52.

26. Koopmans CM, Bijlenga D, Groen H, Vijgen SM, Aarnoudse JG, Bekedam DJ, et al. Induction of labour versus expectant monitoring for gestational hypertension or mild pre-eclampsia after 36 weeks’ gestation (HYPITAT): a multicentre, open-label randomised controlled trial. Lancet. 2009 Sep 19;374(9694):979-88.

27. Tikkanen M, Nuutila M, Hiilesmaa V, Paavonen J, Ylikorkala O. Clinical presentation and risk factors of placental abruption. Acta obstetricia et gynecologica Scandinavica. 2006;85(6):700-5.

28. Toivonen S, Heinonen S, Anttila M, Kosma VM, Saarikoski S. Obstetric prognosis after placental abruption. Fetal diagnosis and therapy. 2004 Jul-Aug;19(4):336-41.

29. Tikkanen M, Nuutila M, Hiilesmaa V, Paavonen J, Ylikorkala O. Prepregnancy risk factors for placental abruption. Acta obstetricia et gynecologica Scandinavica. 2006;85(1):40-4.

30. Yla-Outinen A, Palander M, Heinonen PK. Abruptio placentae--risk factors and outcome of the newborn. European journal of obstetrics, gynecology, and reproductive biology. 1987 May;25(1):23-8.

31. Pritchard JA, Brekken AL. Clinical and laboratory studies on severe abruptio placentae. American journal of obstetrics and gynecology. 1967 Mar 1;97(5):681-700.

32. Hibbard BM, Hibbard ED. Aetiological Factors in Abruptio Placentae. British medical journal. 1963 Dec 7;2(5370):1430-6.

33. Hibbard BM, Jeffcoate TN. Abruptio placentae. Obstetrics and gynecology. 1966 Feb;27(2):155-67.

34. Paterson ME. The aetiology and outcome of abruptio placentae. Acta obstetricia et gynecologica Scandinavica. 1979;58(1):31-5.

35. Rasmussen S, Irgens LM. Occurrence of placental abruption in relatives. BJOG : an international journal of obstetrics and gynaecology. 2009 Apr;116(5):693-9.


97

CHAPTER

6

36. Ananth CV, Savitz DA, Bowes WA, Jr., Luther ER. Influence of hypertensive disorders and cigarette smoking on placental abruption and uterine bleeding during pregnancy. Br J Obstet Gynaecol. 1997 May;104(5):572-8.

37. Surapaneni T, Bada VP, Nirmalan CP. Risk for Recurrence of Pre-eclampsia in the Subsequent Pregnancy. Journal of clinical and diagnostic research : JCDR. 2013 Dec;7(12):2889-91.

38. Hargood JL, Brown MA. Pregnancy-induced hypertension: recurrence rate in second pregnancies. The Medical journal of Australia. 1991 Mar 18;154(6):376-7.

39. Ananth CV, Smulian JC, Vintzileos AM. Incidence of placental abruption in relation to cigarette smoking and hypertensive disorders during pregnancy: a meta-analysis of observational studies. Obstetrics and gynecology. 1999 Apr;93(4):622-8.

CHAPTER 7

Incidence and recurrence rate of severe postpartum

haemorrhage and manual removal of the placenta:

a longitudinal linked national cohort study in

The Netherlands

Ruiter L Kazemier BM

Mol BWJPajkrt E

Submitted

Chapter 7

100

ABSTRACT

Objective We studied the incidence of postpartum haemorrhage (PPH) and manual removal of the placenta and their recurrence rate in a subsequent pregnancy.

Design: Retrospective national cohort study.

Setting: the Netherlands. Population: 359,737 women with two consecutive singleton deliveries between 1999 and 2009.

Methods: A longitudinal linked national cohort with information on subsequent singleton delivery was used.

Main outcome measures: We calculated incidence and recurrence rates of PPH and manual removal of the placenta for all women and stratified by gestational age and mode of delivery.

Results A total of 19,274 (5.4%) women experienced PPH in the first pregnancy and 7536 (2.1%) women had a manual removal of the placenta. In the second pregnancy 16,812 (4.7%) women had PPH and 6,438 (1.8%) women had manual removal of the placenta. The risk of PPH in a subsequent pregnancy was significantly higher in women with a history of PPH compared to women without a previous PPH (18% vs 3.9%, adjusted odds ratio (aOR) 4.5; 95% confidence interval (CI) 4.3-4.7). Compared to women without a history of severe PPH , women PPH and the need for manual removal of the placenta after vaginal or operative vaginal delivery were most at risk for recurrence (aOR 7.5; 95%CI 6.8-8.1 and aOR 8.1; 95%CI 7.2-9.2).

Conclusion Women with a history of PPH after a late preterm or term delivery have an increased risk of PPH in a subsequent pregnancy. The need for a manual removal of the placenta in these women further increases the recurrence rate.

Incidence and recurrence rate of severe postpartum haemorrhage and manual removal of the placenta

101

CHAPTER

7

INTRODUCTION

Postpartum haemorrhage (PPH) is still one of the leading causes of maternal morbidity and mortality in developed and third world countries despite improvements in prevention and management. 1, 2 Recent studies on PPH in developed countries report an increasing incidence over time. 3-5

Among other things, PPH is associated with uterine atony , genital tract lesions and manual removal of the placenta (MRP).6 Risk factors for PPH have been established in several studies and can be divided in antepartum risk factors such as increased maternal age, multiple pregnancy, fibroid tumours, placenta praevia and peripartum risk factors as uterine rupture, prolonged third stage of labor, instrumental vaginal delivery, Cesarean delivery and genital tract lesions. 3, 7, 8 2

Early diagnosis and treatment of threatening PPH during or directly after labor may potentially decrease the incidence of PPH and its associated complications. For example, the administration of oxytocin after delivery of the fetus or placenta has been found effective in the prevention of PPH and is currently standard care in numerous countries, 9 as is an active management in the third stage of labor.10 Potentially, additional individualized care in patients at high risk for PPH can reduce the occurrence and recurrence of this serious complication. In this study we aimed to calculate the incidence and recurrence rate of severe PPH and MRP using record linkage of women’s subsequent singleton pregnancies over a time period between 1999 and 2009. We were particularly interested in the association between characteristics of the first episode of PPH and/or MRP, such as gestational age at first birth and the rate of recurrence in a subsequent pregnancy. Knowledge on this subject can potentially improve counselling, prevention and management in obstetric care for women at risk for PPH.

MATERIALS AND METHODS

We performed a nationwide cohort study using data from the Netherlands Perinatal Registry (PRN). The PRN consists of prospectively collected population-based data containing information on pregnancies, deliveries and neonatal (re)admissions until 28 days after birth. The database is realised by a validated linkage of three separate registries: the midwifery registry (LVR1), the obstetrics registry (LVR2) and the neonatology registry (LNR). 11, 12

Approximately 96% of all deliveries in the Netherlands is documented in the PRN registry. The PRN registry contains all pregnancies ending in live births and still births from 22 weeks of gestation and a minimum birth weight of 500 grams. In addition, all (re-)admissions of newborns to the neonatology care unit are documented. The PRN is primarily designed and used for annual assessment of the quality indicators of obstetric care.

Chapter 7

102

Since all records in the PRN registry are entered at child’s level there is no unique maternal identifier. Consequently, follow-up on outcomes of subsequent pregnancies of the same mother is not possible. For this reason, a mother identifier was created by a longitudinal probabilistic linkage procedure where records of children of the same mother were linked. This resulted in a cohort of 391,026 women with complete data on first and second delivery. Details on this linkage procedure are described elsewhere.13 The data in the PRN database are anonymous and The Netherlands Perinatal Registry gave their approval for the use of their data for this study (approval no. 14.01).

Inclusion and exclusion criteriaWe selected all women that delivered between Jan. 1, 1999 and Dec. 31, 2009. For both pregnancies we excluded women with a multiple gestation and women with a gestational age at delivery <24+0 weeks and > 42+6 weeks. Furthermore, we excluded all women with a placental abruption, since the aetiology of this blood loss is different from the postpartum haemorrhage of interest in our study.

Outcome measuresOur main outcomes were severe postpartum haemorrhage (severe PPH ≥1000mL after vaginal delivery or Cesarean section) and manual removal of the placenta. We selected all women that were registered with a code for severe PPH or MRP in the PRN registry LVR1 and/or LVR2. Diagnosis codes are assigned and registered by midwives, residents and gynaecologists based on the information in the patients’ medical chart after delivery. A diagnosis of severe PPH is assigned with a minimum amount of blood loss of 1000mL, independently of execution of interventions because of PPH. In addition, we labelled all women with a positive history of severe PPH at the beginning of their second pregnancy as cases of severe PPH in the first pregnancy.

CharacteristicsDemographic and obstetric baseline characteristics as ethnicity, socio-economic status (SES), maternal age, fibroid tumours, hypertensive disorders, neonatal sex and birth weight were extracted from the PRN registry. In addition, the following delivery characteristics were extracted: gestational age at delivery in weeks, induction of labor, mode of delivery, fetal presentation at delivery, genital tract lesions and manual placental removal because of retained placental tissue. Ethnicity was categorized into white (native Dutch and other Caucasian) and non-white (different ethnic groups as African/Surinamese, South-Asian, Moroccan and Turkish) for this study. For SES the average household income of subject’s neighbourhood was used based on postal code. Mode of delivery was categorized in spontaneous, primary- and secondary


103

CHAPTER

7

Cesarean section, operative vaginal delivery and vaginal breech delivery. Fetal presentation was cephalic, breech or other. Dating of the pregnancy is based on last menstrual period (LMP) or ultrasound measurements <20weeks of gestation, in case of discrepancy the ultrasound measurement was considered the dominant one.Hypertensive disorders (HTD) included all cases of chronic hypertension, pregnancy-induced hypertension and pre-eclampsia. The clinical diagnosis of hypertension was made with a systolic blood pressure of ≥ 140mmHg and/or a diastolic blood pressure of ≥ 90mmHg. Pre-eclampsia was diagnosed with hypertension and proteinuria.14 Cases of severe PPH and MRP were divided in subgroups; 1 women with severe PPH and MRP, 2 women with severe PPH without MRP and 3 women with only MRP and were compared to women without a history of both severe PPH or MRP. To estimate the effect of gestational age at delivery in the first pregnancy women with and without a history of severe PPH (irrespectively of MRP) were compared after stratification by first early preterm (24+0 - 31+6 weeks gestation), late preterm (32+0 - 36+6 weeks gestation) and term (37+0 - 42+6 weeks gestation) delivery. In addition, we stratified the analysis by characteristics of the first delivery namely mode of delivery and need for manual removal of the placenta, comparing women with and without a history of severe PPH.

AnalysisThe incidences of severe PPH and MRP were calculated for first and second pregnancies by dividing the number of deliveries complicated by severe PPH, the total number of MRP was divided by the total number of vaginal deliveries. Calculated frequencies represent the population risk for severe PPH and MRP respectively. In addition, we calculated the incidences of severe PPH and MRP per gestational week by dividing all cases by the total number of (vaginal for MRP) deliveries in the corresponding week.

We compared demographic and first pregnancy and delivery baseline characteristics between the four subgroups of women. Normally distributed continuous variables were compared by univariate analysis with the ANOVA test and categorical variables were compared with the chi-square test. All tests were two-sided with a p-value of 0.05 as the threshold for statistical significance. To estimate the effect of a previous severe PPH and/or MRP on the recurrence of these complications in a subsequent pregnancy logistic regression modelling was used and was expressed as odds ratio (OR) with a 95% confidence interval (CI). To account for the effect of potential confounders we used a multivariate logistic regression analysis to adjust for variables that were unequally distributed in the baseline characteristics of the first pregnancy The effect of gestational age at first occurrence of severe PPH on the recurrence rate was

Chapter 7

104

Tabl

e 1.

Bas

elin

e ch

arac

teris

tics o

f the

com

plet

e lin

ked

coho

rt (N

=359

,737

)

Firs

t pre

gnan

cy

Mat

erna

l cha

ract

eris

tics

No

PPH,

no

MRP

(n=

337

,229

)PP

H w

ith M

RP

(n=

4,3

02)

PPH

with

out M

RP

(n=

14,9

72)

MRP

with

out P

PH

(n=

3,23

4)p-

valu

e

n%

n%

n%

n%

Mat

erna

l age

, yea

rs ±

SD

28.5

(± 4

.2)

30.2

(± 3

.7)

28.8

(± 4

.1)

29.8

(± 4

.0)

<0,0

001

Mat

erna

l age

<25

y, n

(%)

5749

017

,0%

303

7,0%

2099

14,0

%31

19,

6%<0

,000

1M

ater

nal a

ge ≥

35 y,

n (%

)23

843

7,1%

514

11,9

%10

937,

3%36

211

,2%

<0,0

001

Non

-whi

te m

ater

nal e

thni

city

, n (%

)39

270

11,6

%21

45,

0%14

9210

,0%

253

7,8%

<0,0

001

Low

soc

ioec

onom

ic st

atus

, n (%

)75

301

22,3

%73

917

,2%

3065

20,5

%65

220

,2%

<0,0

001

Hyp

erte

nsiv

e di

sord

er, n

(%)

4100

112

,2%

686

15,9

%21

3414

,3%

609

18,8

%<0

,000

1Fi

broi

d tu

mor

s63

70,

2%20

0,5%

510,

3%5

0,2%

<0,0

001

Deliv

ery

char

acte

ristic

sM

ean

gest

ation

al a

ge, w

k ±

SD39

w1d

(±

2w0d

)39

w1d

(± 2

w4d

)39

w4d

(± 2

w0d

)38

w1d

(± 3

w6d

)<0

,000

1Pr

emat

ure

birt

h <

37w

k24

544

7,3%

473

11,0

%78

65,

2%65

920

,4%

<0,0

001

Spon

tane

ous

star

t of l

abor

, n (%

)19

5284

57,9

%18

8543

,8%

8052

53,8

%13

0340

,3%

<0,0

001

Mod

e of

del

iver

y<0

,000

1Sp

onta

neou

sly21

4213

63,5

%28

9667

,3%

9135

61,0

%19

1759

,3%

Prim

ary

Cesa

rean

sec

tion

1844

45,

5%8

0,2%

590

3,9%

146

4,5%

Seco

ndar

y Ce

sare

an s

ectio

n40

216

11,9

%11

0,3%

1566

10,5

%15

94,

9%O

pera

tive

vagi

nal d

eliv

ery

6412

619

,0%

1381

32,1

%36

7324

,5%

997

30,8

%Br

eech

del

iver

y23

00,

1%6

0,1%

80,

1%15

0,5%

Perin

eal o

r lab

ial t

ear*

1034

5437

,1%

1372

32,0

%46

5936

,4%

828

28,3

%<0

,000

1Ep

isiot

omy*

1436

8751

,6%

2644

61,7

%78

4661

,2%

1650

56,3

%<0

,000

1Fe

tal p

rese

ntati

on<0

,000

1Ce

phal

ic30

8185

91,4

%41

6396

,8%

1433

795

,8%

2914

90,1

%Br

eech

2412

57,

2%11

22,

6%47

83,

2%27

58,

5%O

ther

4919

1,5%

270,

6%15

71,

0%45

1,4%

Feta

l cha

ract

eris

tics

Mal

e, n

(%)

1736

6951

,5%

2049

47,6

%73

3549

,0%

1623

50,2

%<0

,000

1M

ean

birt

hwei

ght (

gram

s ±

SD)

3358

(± 5

82)

3369

(± 6

49)

3576

(± 5

79)

3056

(± 8

79)

<0,0

001

Mac

roso

mia

>40

0038

333

11,4

%61

914

,4%

3253

21,7

%26

48,

2%<0

,000

1M

acro

som

ia >

4500

5514

1,6%

121

2,8%

623

4,2%

401,

2%<0

,000

1

* %

of v

agin

al d

eliv

erie

s

PPH

pos

tpar

tum

hem

orrh

age

M

RP m

anua

l rem

oval

of p

lace

nta


105

CHAPTER

7

estimated and expressed as odds ratio with 95% CI, comparing women with a history of severe PPH (both with or without MRP) to women without severe PPH or MRP in the first pregnancy.

To estimate the effect of mode of delivery prior to the first episode of severe PPH and the effect of manual placental removal (MPR), multivariate logistic regression modelling was used comparing the subgroups of mode of delivery to women without a history of severe PPH and MRP. Data were analysed with the SAS statistical software package (version 9.4; SAS Institute Inc, Cary, NC, USA).

RESULTS

Between Jan. 1, 1999 and Dec. 31, 2009, a total of 359,737 singleton-singleton pregnancies met all inclusion criteria and were included in our study. A total number of 10,800 (2.8%) multiple gestations, 2,754 (0.7%) pregnancies with a gestational age at delivery <24+0 weeks and > 42+6 weeks and 335 (0.1%) pregnancies complicated by a placental abruption were excluded. The baseline characteristics of our linked cohort are listed in Table 1. Women with a severe PPH and/or MRP after their first delivery were older, more often of white ethnicity and had more often hypertensive disorders compared to women without severe PPH or MRP in their first pregnancy (P<0.0001). Operative vaginal delivery, genital tract lesions because of episiotomy and macrosomia were also more prevalent in women with severe PPH and/or MRP compared to the non-PPH and non-MRP group (P<0.0001).A total of 19,274 (5.4%) women had severe PPH in their first pregnancy. The total number of MRP was 7,536 (2.1%) in the first pregnancy. The overall incidences of severe PPH (with and without MRP) and MRP in the first pregnancy per gestational week are shown in supplemental figures S1a and S1b. Of 359,737 first singleton deliveries, 4,302 (1.2%) were complicated by severe PPH and MRP, 14,972 (4.2%) were complicated by severe PPH without MRP and 3,234 (0.90%) were complicated by MRP only. Of 359,737 second singleton deliveries, 4,132 (1.1%) were complicated by severe PPH and MRP, 12,680 (3.5%) were complicated by severe PPH without MRP and 2,306 (0.64%) were complicated by MRP only. The total number of severe PPH in the second pregnancy was 16,812 (4.7%), the total number of MRP was 6,438 (1.8%). The recurrence rates of severe PPH and/or MRP are shown in table 2, as well as the corresponding odds ratio of a history of severe PPH and/or MRP on the occurrence of these complications in the second pregnancy. In the multivariate regression analysis adjustment was made for maternal age, ethnicity, hypertension, gestational age, induction of labor, genital tract lesions, fetal presentation and sex and macrosomia >4500g.

Chapter 7

106

Tabl

e 2a

and

2b.

Per

cent

ages

of r

ecur

renc

e an

d ad

just

ed o

dds r

atios

2a. P

erce

ntag

es o

f rec

urre

nce

Seco

nd d

eliv

ery

Char

acte

ristic

s of

firs

t del

iver

y:N

o PP

H, n

o M

RP

(n=3

40,6

19)

PPH

and

MRP

(n

=4,1

32)

PPH

with

out M

RP

(n=1

2,68

0)M

RP w

ithou

t PPH

(n=2

,306

)n

%n

%n

%n

%

No

PPH,

no

MRP

(n=3

37,2

29)

3224

7696

%27

610.

8%10

181

3.0%

1811

0.5%

PPH

and

MRP

(n=4

,302

)30

6671

.3%

586

13.6

%47

211

.0%

178

4.1%

PPH

with

out M

RP (n

=14,

972)

1237

882

.7%

582

3.9%

1856

12.4

%15

61.

0%M

RP w

ithou

t PPH

(n=3

,234

)26

9983

.5%

203

6.3%

171

5.3%

161

5.0%

2b. A

djus

ted

odds

ratio

sA

Seco

nd d

eliv

ery

Char

acte

ristic

s of

firs

t del

iver

y:N

o PP

H, n

o M

RP

(n=3

40,6

19)

PPH

and

MRP

(n

=4,1

32)

PPH

with

out M

RP

(n=1

2,68

0)M

RP w

ithou

t PPH

(n

=2,3

06)

No

PPH

, no

MRP

(n=3

37,2

29)

Refe

renc

eRe

fere

nce

Refe

renc

eRe

fere

nce

PPH

and

MRP

(n=4

,302

)0.

12 (0

.11-

0.13

)18

(16-

19)

3.8

(3.5

-4.2

)7.

3 (6

.2-8

.6)

PPH

with

out M

RP (n

=14,

972)

0.22

(0.2

1-0.

23)

4.8

(4.4

-5.3

)4.

4 (4

.2-4

.7)

1.9

(1.6

-2.3

)M

RP w

ithou

t PPH

(n=3

,234

)0.

24 (0

.22-

0.27

)7.

5 (6

.5-8

.7)

1.8

(1.5

-2.0

)8.

6 (7

.2-1

0)

PPH

pos

tpar

tum

hae

mor

rhag

e

MRP

man

ual r

emov

al o

f pla

cent

aA

Adju

sted

for m

ater

nal a

ge, e

thni

city

, hyp

erte

nsio

n, g

esta

tiona

l age

, ind

uctio

n of

labo

r, ge

nita

l tra

ct le

sions

, fet

al p

rese

ntati

on, f

etal

sex

, mac

roso

mia


107

CHAPTER

7

Table 3. Severe PPH in a subsequent pregnancy in total and by gestational age of delivery in the first pregnancy

First pregnancy PPH second pregnancy Odds ratio (95% CI)Total n % Crude Adjusted

Women without PPH 340463 13316 3.9% Reference

Women with PPH 19274 3496 18.1% 4.9 (4.6-5.1) 4.5 (4.3-4.7)A

Gestational age at delivery:<32 weeks 208 16 7.7% 2.1 (1.2-3.4) 1.4 (0.84-2.4)B

32-37 weeks 1051 183 17.4% 5.2 (4.4-6.1) 4.0 (3.4-4.7)B

>37 weeks 18015 3297 18.3% 5.5 (5.3-5.7) 4.6 (4.4-4.8)B

A Adjusted for manual placental removal, maternal age, ethnicity, hypertension, gestational age, induction of labor, genital tract lesions, fetal presentation, fetal sex, macrosomia)B Adjusted for manual placental removal, maternal age, ethnicity, hypertension, induction of labor, genital tract lesions, fetal presentation, fetal sex and macrosomia (>4500g)

Table 4. Recurrence of severe PPH by mode of delivery and need for manual placental removal in the first pregnancy

First pregnancy PPH second pregnancy Odds ratio (95% CI)

Total n % Crude Adjusted

Women without PPH 340463 13316 3.9% Reference Reference

Women with PPH 19274 3496 18.1%Vaginal delivery with MRP 2896 682 23.5% 7.6 (6.9-8.3) 7.5 (6.8-8.1)Vaginal delivery without MRP 9135 1557 17.0% 5.0 (4.8-5.3) 5.3 (5.0-5.6)Operative vaginal delivery with MRP 1387 372 26.8% 9.0 (8.0-10) 8.1 (7.2-9.2)Operative vaginal delivery without MRP 3681 603 16.4% 4.8 (4.4-5.3) 4.5 (4.1-4.9)Primary Cesarean section 598 91 15.2% 4.4 (3.5-5.5) 3.4 (2.7-4.2)Secondary Cesarean section 1577 191 12.1% 3.4 (2.9-3.9) 2.9 (2.5-3.4)

Women with a history of severe PPH and MRP were at increased risk of recurrence of severe PPH and MRP in the second pregnancy (aOR 18; 95%CI 16-19), as well as for severe PPH without MRP (aOR 3.8; 95%CI 3.5-4.2) and MRP only (aOR 7.3; 95%CI 6.2-8.6) compared to women without a history of either severe PPH or MRP. Women with a history of severe PPH without MRP were also at increased risk of recurrence of severe PPH without MRP in the second pregnancy (aOR 4.4; 95%CI 4.2-4.7), as well as for severe PPH with MRP (aOR 4.8; 95%CI 4.4-5.3) and MRP only (aOR 1.9; 95%CI 1.6-2.3) compared to women without a history of either severe PPH or MRP.

Chapter 7

108

Women with a history of MRP without severe PPH were at increased risk of recurrence of this complication in the second pregnancy (aOR 8.6; 7.2-10), as well as for MRP with severe PPH (aOR 7.5; 95%CI 6.5-8.7) and severe PPH without MRP (aOR 1.8; 95%CI 1.5-2.0) compared to women without a history of either severe PPH or MRP. Table 3 shows the gestational age at first delivery complicated by severe PPH (both with and without MRP) and the corresponding recurrence rate in a subsequent pregnancy. The recurrence rate was increased for women with a history of severe PPH following a delivery in the term (aOR 4.6; 95% CI 4.4–4.8) and late preterm (aOR 4.0; 95% CI 3.4-4.7) period of pregnancy compared to women without a history of severe PPH. A history of first delivery

Figure 1. Incidence of postpartum hemorrhage by gestational age per 1000 deliveries

Figure 2. Incidence of manual placental removal by gestational age per 1000 deliveries


109

CHAPTER

7

in the early preterm period of pregnancy complicated by severe PPH did not significantly increase the risk of severe PPH in a subsequent pregnancy compared to women without a history of severe PPH (aOR 1.4; 95%CI 0.8-2.4). Table 4 shows that women with a history of severe PPH following a vaginal (aOR 7.5; 95%CI 6.8-8.1) or instrumental vaginal (aOR 8.1; 95%CI 7.2-9.2) delivery with need for MRP were most at risk for recurrence of severe PPH in a subsequent pregnancy compared to women without a history of severe PPH.

COMMENTS

In this linked cohort study we investigated the incidence and recurrence rate of severe PPH and MRP among singleton pregnancies in the Netherlands. In addition, we investigated the effect of gestational age and mode of first delivery on the recurrence rate.

Main findingsThe incidence of severe PPH in first pregnancies was 5.4% and the incidence increases with increasing gestational age. The incidence of MRP was 2.1% in first pregnancies and the incidence decreases with increasing gestational age. The risk of severe PPH in a subsequent pregnancy was significantly higher in women with a previous severe PPH compared to women without a history of severe PPH, the same applies to MRP.Importantly, this study shows that the recurrence rate of severe PPH in a subsequent pregnancy is associated with the gestational age of the first delivery that was complicated by severe PPH. Women with a history of severe PPH after a term delivery or late preterm delivery are at risk of recurrence compared to women without a history of severe PPH. A history of severe PPH after a delivery of an early preterm pregnancy was not associated with an increased recurrence rate in a subsequent pregnancy.

Strengths and limitationsThis study comes with both strengths and limitations. One of the strengths of this study is that it is performed with data from a large national registry with prospectively collected data thereby minimizing recall and reporting bias. We needed to rely on the information that was available in the perinatal registry thus we cannot rule out misclassification and potential under registration. However, since a history of severe PPH or MRP is a well-known indication for hospital delivery in a subsequent pregnancy in the Netherlands we do not expect that a significant amount of cases of severe PPH and MRP have not been registered. Furthermore, data on possible confounders and potential risk factors for severe PPH as body mass index , tobacco use and bleeding disorders are unavailable. In addition, the total amount of blood

Chapter 7

110

loss is not registered in the PRN registry which made it impossible to account for the degree of severity of PPH.

InterpretationThe incidence of severe PPH in literature varies between 2.3 and 10%. But since the definition of PPH is not uniform and the incidence of PPH appears to follow an increasing trend over time, exact comparison is difficult. Numerous studies used a total amount of blood loss of 500mL as cut-off for PPH whereas in the Netherlands a cut- off of 1000mL is used. Since blood loss of this severity rarely comprises maternal condition we think our definition is clinically more relevant.8, 15 When we compare our data to studies with data from a similar study period and definition (we find comparable incidences, varying between 5.1 and 6.8%. 4, 5, 16 The incidence in our study is slightly higher than the incidence calculated in studies based on data before the year 2000, which was on average 4.0%. 5, 17 This is in line with the increasing trend on PPH incidence described in previous studies. 3, 4 18 The incidence of MRP in our study is comparable to the incidences reported in literature, between 0.5 and 3.0%.6, 19-23 Whereas the incidence of PPH increased over time, the incidence of MRP remained relatively constant over the last decades. In our study we found a recurrence rate of PPH in a subsequent pregnancy of 18.1% which is comparable with previous documented recurrence rates. 5, 18 Potentially, the recurrence rate is even higher than reported since it is known that women with a history of severe PPH are less likely to conceive again.17 Knowledge on a history of PPH in prenatal care may alter the obstetric care for women that do conceive again. A previous study of Oberg et al. found that the risk of recurrence of PPH was significantly increased nonetheless the cause of the first episode of PPH, however the total increase in risk may be different. They found that women with a vaginal delivery complicated by PPH because of retained placental tissue were more at risk for recurrence of PPH than women with PPH because of atony or genital tract lacerations, which is in line with our results that women with PPH and MRP are more at risk for recurrence than women with PPH without MRP.18 Although we were unable to stratify PPH by subtypes in our study due to the unavailability of this information in the PRN, we did find that women with a history of spontaneous vaginal or operative vaginal delivery with PPH and manual removal of the placenta were more at risk for recurrence than women without a history of manual placental removal, partially supporting this conclusion of Oberg et al. In our study we extended the findings of previous studies on the recurrence rate by investigating the effect of gestational age at delivery at first episode of PPH. Although a higher incidence of PPH in term pregnancies beyond 40 weeks of gestation has been reported in literature before, comparison with preterm deliveries nor the recurrence rate has been investigated before.24 To our knowledge, our study is the first to stratify recurrence rate of PPH by gestational age of the first episode of PPH. Women with a late preterm or term delivery


111

CHAPTER

7

complicated by PPH are at increased risk of occurrence of PPH in a subsequent pregnancy compared to women without a history of PPH. However, women with an early preterm delivery complicated by PPH were not significantly at increased risk for occurrence of PPH in a subsequent pregnancy compared to women without a history of PPH. We hypothesize that the limited development of preterm myometrial tissue rather than dysfunctional myometrial tissue may explain why cases of preterm PPH recur less in a subsequent pregnancy than cases of term PPH compared to women without a history of PPH.19 The latter may be more prone to recur because of an unknown dysfunctional pathway after birth of this myometrial tissue. Thus, although preterm delivery is known to increase the initial risk of PPH and MRP 19, 21, 25, 26, the consequences for a potential future pregnancy and delivery seem to be less relevant clinically . Women with a history of early preterm delivery complicated by PPH can thus potentially be reassured when considering conceiving again.

CONCLUSION

In conclusion, the findings of this study show that women with a previous delivery >32 weeks complicated by PPH are at increased risk for recurrence especially when there was a need for manual placental removal compared to women without a history of PPH. Longitudinal linked cohort studies with complete information on type of PPH, gestational age at delivery and total amount of blood loss are needed to develop prediction models to further stratify low- and high risk women for PPH. Women at high risk of PPH and/or MRP should be delivered in a setting with adequate obstetric and anaesthetic interventions and personnel. Stratification and risk assessment can help in counselling and individualization of prenatal and peripartum care in women with a history of PPH to potentially decrease the recurrence of this serious complication.

Chapter 7

112

REFERENCES

1. Mathai M, Gulmezoglu AM, Hill S. Saving womens lives: evidence-based recommendations for the prevention of postpartum haemorrhage. Bulletin of the World Health Organization. 2007 Apr;85(4):322-3.

2. World Health Organization. WHO Recommendations for the prevention and treatment of postpartum haemorrhage. . Geneva (Switzerland); 2012.

3. Kramer MS, Berg C, Abenhaim H, Dahhou M, Rouleau J, Mehrabadi A, et al. Incidence, risk factors, and temporal trends in severe postpartum hemorrhage. American journal of obstetrics and gynecology. 2013 Nov;209(5):449.e1-7.

4. Joseph KS, Rouleau J, Kramer MS, Young DC, Liston RM, Baskett TF. Investigation of an increase in postpartum haemorrhage in Canada. BJOG : an international journal of obstetrics and gynaecology. 2007 Jun;114(6):751-9.

5. Ford JB, Roberts CL, Bell JC, Algert CS, Morris JM. Postpartum haemorrhage occurrence and recurrence: a population-based study. The Medical journal of Australia. 2007 Oct 1;187(7):391-3.

6. Hall MH, Halliwell R, Carr-Hill R. Concomitant and repeated happenings of complications of the third stage of labour. Br J Obstet Gynaecol. 1985 Jul;92(7):732-8.

7. Combs CA, Murphy EL, Laros RK, Jr. Factors associated with postpartum hemorrhage with vaginal birth. Obstetrics and gynecology. 1991 Jan;77(1):69-76.

8. Ramanathan G, Arulkumaran S. Postpartum haemorrhage. Current Obstetrics & Gynaecology. 2006;16(1):6-13.

9. Westhoff G, Cotter AM, Tolosa JE. Prophylactic oxytocin for the third stage of labour to prevent postpartum haemorrhage. The Cochrane database of systematic reviews. 2013;10:Cd001808.

10. Prendiville WJ, Elbourne D, McDonald S. Active versus expectant management in the third stage of labour. The Cochrane database of systematic reviews. 2000(3):CD000007.

11. Meray N, Reitsma JB, Ravelli AC, Bonsel GJ. Probabilistic record linkage is a valid and transparent tool to combine databases without a patient identification number. Journal of clinical epidemiology. 2007 Sep;60(9):883-91.

12. Tromp M, Ravelli AC, Meray N, Reitsma JB, Bonsel GJ. An efficient validation method of probabilistic record linkage including readmissions and twins. Methods of information in medicine. 2008;47(4):356-63.

13. Schaaf JM, Hof MH, Mol BW, Abu-Hanna A, Ravelli AC. Recurrence risk of preterm birth in subsequent singleton pregnancy after preterm twin delivery. American journal of obstetrics and gynecology. 2012 Oct;207(4):279 e1-7.

14. Guideline Hypertensive disorders in pregnancy. Utrecht: The Dutch Society of Obstetrics and Gynecology; 2011.

15. Lu MC, Korst LM, Fridman M, Muthengi E, Gregory KD. Identifying women most likely to benefit from prevention strategies for postpartum hemorrhage. Journal of perinatology : official journal of the California Perinatal Association. 2009 Jun;29(6):422-7.

16. Magann EF, Evans S, Hutchinson M, Collins R, Howard BC, Morrison JC. Postpartum hemorrhage after vaginal birth: an analysis of risk factors. Southern medical journal. 2005 Apr;98(4):419-22.

17. Sentilhes L, Gromez A, Clavier E, Resch B, Descamps P, Marpeau L. Long-term psychological impact of severe postpartum hemorrhage. Acta obstetricia et gynecologica Scandinavica. 2011 Jun;90(6):615-20.

18. Oberg AS, Hernandez-Diaz S, Palmsten K, Almqvist C, Bateman BT. Patterns of recurrence of postpartum hemorrhage in a large population-based cohort. American journal of obstetrics and gynecology. 2014 Mar;210(3):229.e1-8.


113

CHAPTER

7

19. Coviello EM, Grantz KL, Huang CC, Kelly TE, Landy HJ. Risk factors for retained placenta. American journal of obstetrics and gynecology. 2015 Dec;213(6):864 e1- e11.

20. Endler M, Saltvedt S, Cnattingius S, Stephansson O, Wikstrom AK. Retained placenta is associated with pre-eclampsia, stillbirth, giving birth to a small-for-gestational-age infant, and spontaneous preterm birth: a national register-based study. BJOG : an international journal of obstetrics and gynaecology. 2014 Nov;121(12):1462-70.

21. Endler M, Grunewald C, Saltvedt S. Epidemiology of retained placenta: oxytocin as an independent risk factor. Obstetrics and gynecology. 2012 Apr;119(4):801-9.

22. Owolabi AT, Dare FO, Fasubaa OB, Ogunlola IO, Kuti O, Bisiriyu LA. Risk factors for retained placenta in southwestern Nigeria. Singapore Med J. 2008 Jul;49(7):532-7.

23. Cheung WM, Hawkes A, Ibish S, Weeks AD. The retained placenta: historical and geographical rate variations. J Obstet Gynaecol. 2011;31(1):37-42.

24. Caughey AB, Bishop JT. Maternal complications of pregnancy increase beyond 40 weeks of gestation in low-risk women. Journal of perinatology : official journal of the California Perinatal Association. 2006 Sep;26(9):540-5.

25. Adelusi B, Soltan MH, Chowdhury N, Kangave D. Risk of retained placenta: multivariate approach. Acta obstetricia et gynecologica Scandinavica. 1997 May;76(5):414-8.

26. Romero R, Hsu YC, Athanassiadis AP, Hagay Z, Avila C, Nores J, et al. Preterm delivery: a risk factor for retained placenta. American journal of obstetrics and gynecology. 1990 Sep;163(3):823-5.

CHAPTER 8

Summary and general discussion


117

CHAPTER

8

SUMMARY

This thesis focused on vasa previa and pregnancy complications associated with abnormalities of the placenta.Vasa previa on ultrasound was first reported in 1987 and has received increasingly attention over the last decades.1 Since the incidence of vasa previa is relatively low, literature on this complication is mostly limited to case reports, case series and reviews. The majority of the studies emphasize the importance of prenatal diagnosis because of the potential fatal neonatal outcome in pregnancies complicated by vasa previa. In many countries including the Netherlands, clear guidelines regarding prenatal diagnosis for vasa previa were unavailable. Therefore, in the first part of this thesis we focused on vasa previa. We evaluated the diagnostic accuracy of ultrasound in the prenatal detection of vasa previa and we investigated the incidence of and identified potential risk indicators for vasa previa. As a final step we initiated the implementation of a national guideline regarding prenatal screening and management in vasa previa.

As vasa previa, other abnormalities of the placenta such as placenta previa, placental abruption and retained placenta also have been associated with adverse maternal and neonatal outcomes. Placenta previa is a condition commonly seen during pregnancy, but the clinical presentation varies between asymptomatic and substantial vaginal blood loss, impeding the efficacy of an uniform guideline. The second part of this thesis aimed to differentiate between high-risk and low-risk women with placenta previa, providing opportunities for individualized antenatal management. Furthermore, we investigated the consequences of placental abruption and retained placental tissue with or without postpartum hemorrhage on a subsequent pregnancy to increase individualized prenatal care.

In this final chapter we will summarize the main findings of our work and outline them within the context of the current literature. Furthermore we will provide implications for further clinical practice and research.

Prenatal diagnosis and management of vasa previaIn current clinical practice multiple sonographic scans are performed during pregnancy, including the second trimester anomalies scan. The possibility to detect vasa previa prenatally has been described in many studies before, though the diagnostic performance of ultrasound in the prenatal detection of vasa previa was unknown.2-4 In chapter two we performed a systematic review on the accuracy of ultrasound. We concluded that the accuracy of ultrasound is high, if performed transvaginally and combined with color Doppler. We found prenatal detection rates between 53% and 100%. The sensitivity was 100% and the specificity

Chapter 8

118

was between 99 and 100%. The diagnosis of vasa previa was best made between 18 and 26 weeks of gestation. An increase in missed cases was seen when scans were performed only trans abdominal and/or exclusively in the third trimester.

Based on the conclusions of publications regarding vasa previa, it appears to be a significant obstetric complication and several recommendations for the implementation of a screening program are made. Unfortunately, the majority of publications are case reports and case series limited by their level of evidence. To objectively estimate the significance of vasa previa, in chapter three we performed a systematic review and meta-analysis on the incidence of vasa previa. Furthermore we identified risk indicators for vasa previa. We found vasa previa to have a mean incidence of 0.60 per 1000 pregnancies or 1 in 1667 pregnancies. Risk indicators for vasa previa were placenta previa in the second trimester, velamentous cord insertion, conception by assisted reproductive technologies, placental morphological anomalies such as bilobed or succenturiate placenta and lower umbilical cord insertion at first trimester ultrasound.

In chapter four we combined our results on vasa previa found in the previous chapters. In addition, we searched the literature for evidence on the management of vasa previa. The combination of these resulted in a national guideline which will be implicated in clinical practice in the Netherlands and advises to actively search for vasa previa in the presence of risk indicators.

Placenta previa and the risk for emergency Cesarean deliveryThe current state of the art in women with diagnosed with placenta previa by ultrasound in the second trimester is to perform a follow-up scan in the third trimester. In case of a persistent placenta previa, an elective Cesarean delivery is scheduled. Several recommendations for timing of delivery are available and mainly concentrate on the late preterm period. Because of prematurity, administration of antenatal corticosteroids is suggested. But some pregnancies with placenta previa remain uncomplicated and are in no need for premature delivery, therefore it is important to distinguish high from low risk women to minimize iatrogenic prematurity. In a national cohort described in chapter five we found a history of Cesarean section, increasing number of episodes of antepartum bleeding and need for blood transfusion as independent predictors for an emergency delivery. In this cohort 43% of the women delivered in an emergent setting at a mean gestational age of 35weeks and 3 days, the majority of these women had one or more episodes of antenatal bleeding. On the contrary, elective Cesarean deliveries in this Dutch cohort were scheduled from 38weeks of gestation and 57% of the women reached this date. This supports the potential reduction of iatrogenic preterm delivery in women with placenta previa, stratification thus remains very important.


119

CHAPTER

8

Consequences for a subsequent pregnancy after placental abruption or retained placental tissueIt is long known that placental abruption is associated with neonatal mortality and morbidity. Several risk factors have been identified including a history of placental abruption although most studies on this subject included relatively small samples and are dated.5 In a large retrospective national cohort study in chapter six we found an incidence of placental abruption of 0.22% among 1,570,635 singleton pregnancies. The incidence was significantly higher in women with a hypertensive disorder compared to normotensive women. The risk of placental abruption in a subsequent pregnancy was 5.8% vs 0.06% for women with and without a history of placental abruption. We found the risk of recurrence to be associated with the gestational age of first placental abruption. Women with a placental abruption in the term period of their first pregnancy were more at risk of recurrence than women with an early preterm (<32weeks) or late preterm (32-37weeks) placental abruption in their first pregnancy compared to women without a history of placental abruption. Although hypertensive disorders increased the risk of placental abruption, no interaction was found on the recurrence rate.

In chapter seven we investigated the incidences of severe postpartum hemorrhage (>1000mL) and retained placental tissue in a cohort of 359,737 women. Retained placental tissue with and without severe postpartum hemorrhage occurred in 2.1% and 5.4% respectively in the first pregnancy. For women with a history of severe postpartum hemorrhage the risk in a subsequent pregnancy was significantly higher compared to women without a history of severe postpartum hemorrhage (18% vs. 3.9%). A history of retained placental tissue followed by severe postpartum hemorrhage increased the risk of recurrence even more compared to women with a history negative for these complications. We found that gestational age of the first delivery complicated by severe postpartum hemorrhage is associated with the rate of recurrence. Women with a history of severe postpartum hemorrhage after a term or late preterm delivery were at risk of recurrence compared to women without a history of severe postpartum hemorrhage. On the contrary, women with a history of an early preterm delivery before 32weeks did not have a significantly increased risk of recurrence of severe postpartum hemorrhage.

Chapter 8

120

GENERAL DISCUSSION

Conclusion, clinical implications and suggestions for future researchSince vasa previa complicates one in 1667 pregnancies, can be accurately diagnosed by ultrasound in pregnancy and can result in neonatal mortality we should aim to optimize prenatal management. The diagnosis of vasa previa should be made by experienced clinicians because either false negative and false positive diagnoses imply several neonatal and maternal consequences. A false negative diagnosis may result in rupture of membranes and one or more fetal vessels in a setting outside the hospital where immediate delivery by Cesarean section and sufficient resuscitation of the neonate cannot be realized. On the contrary, a false positive diagnosis creates uncertainty and stress among the future parents and moreover it results in iatrogenic preterm Cesarean delivery with respiratory problems in het neonate and maternal consequences for potential future pregnancies. Because of the potential serious consequences associated with vasa previa any suspicion of this complication during ultrasound legitimizes referral to experienced clinicians.

As the majority of vasa previa is preceded by one or more risk factors, we do not advise a screening strategy as part of routine midgestation scanning in the general population. The publication of Cipriano et al supports this by demonstrating that general screening for vasa previa is not cost-effective.6 In this thesis we suggest to exclude vasa previa in women with one or more risk indicators. Therefore, one should actively search for velamentous cord insertion, bilobed or succenturiate placenta and placenta previa at the midgestation scan and perform a transvaginal color Doppler in one of these situations as defined in chapter four.

For the Netherlands, implementation of a screening program involves additional demands on the prenatal care provided in and outside hospitals nowadays, including additional training in sonography. Current midgestation scanning should be augmented with color Doppler. Besides the acquirement of practical skills, there is a need to increase knowledge and awareness regarding vasa previa. Previous research found one third of questioned obstetricians not sufficiently expert to name at least one risk factor for vasa previa.7

Besides the implementation of a screening program, we must contemplate the perinatal management in case of vasa previa. It is presumed that ruptured vasa previa almost certainly ends in perinatal death but the effectiveness of a Cesarean delivery to prevent hemorrhage of the fetal vessels is never investigated. Randomized clinical trials regarding the method of delivery are not possible due to the low incidence, but the extent of neonatal morbidity and mortality due to vasa previa can potentially be investigated by evaluation of all cases of perinatal death due to exsanguination and all neonates that required blood transfusion over a certain period of time.


121

CHAPTER

8

Consensus based guidelines advise elective Cesarean delivery from 35 weeks of gestation. It is believed that the mild complications associated with prematurity outweigh the complications that could result from spontaneous rupture of membranes in women with vasa previa. A decision analysis is available on the effectiveness of different timing strategies for delivery in vasa previa, varying gestational age and administration of corticosteroids. Since Cesarean delivery was performed for all individuals, strategies were based on differences in perinatal but not maternal outcomes.6 This decision analysis indicates that the preferred timing of delivery is 34 or 35 weeks of gestation under most circumstances. Under no circumstances a benefit was to be expected from delivery beyond 37 weeks of gestation. However, this study does not allow for obstetric history. Women with a previous preterm birth <34weeks of gestation are at risk for recurrence of preterm birth and should potentially receive more intensive monitoring and earlier delivery in case of vasa previa than nulliparous women or women with a previous term delivery. In addition to timing of delivery many questions remain unanswered. Should women with a prenatal diagnosis of vasa previa be admitted to the hospital prior to delivery and receive corticosteroids? Does hospital admittance reduce the rates of fetal exsanguination and fetal death in spontaneous rupture of membranes in case of vasa previa? An alternative could be to monitor cervical length (and additional fetal fibronectin test in case of shortening) in women with vasa previa in an outpatient setting. In a retrospective cohort study by Baulies et al women remained at home with instructions to attend to the clinic in case of vaginal bleeding or contractions, all neonates were born in a clinical good condition at a mean gestational age of 36 weeks.3 Since prospective clinical trials are likely to be precluded, results from cohort studies and decision analyses might be the only way to answer the question on perinatal management in vasa previa. With the information available from these studies, we should strive to nationalize our guideline and increase awareness on the subject of vasa previa.

Clinical implications that can be made on the second part of this thesis can be summarized as follows. Women with a diagnosis of placenta previa in the second trimester should be followed up in the third trimester, as many cases of placenta previa resolve over time during pregnany. How many of these placenta previas resolve over time is unknown; at this moment we are conducting a prospective cohort study in the Academic Medical Center to evaluate all placenta localizations from the second trimester of pregnancy. However, in case the placenta is no longer low lying in the third trimester, an additional transvaginal color Doppler should be performed to exclude vasa previa. In case of persistent placenta previa, perinatal management in women diagnosed with placenta previa should become more individualized. Women with asymptomatic placenta previa and absence of a

Chapter 8

122

Cesarean delivery in their obstetric history can be monitored in an outpatient setting and scheduled for elective Cesarean delivery at 37 or 38 weeks of gestation.8, 9 Administration of corticosteroids can then be omitted. However, in case of contractions and/or (recurrent) vaginal bleeding women should be admitted to the hospital and scheduled for Cesarean delivery from 35-36 weeks of gestation after administration of corticosteroids, as the chance for emergency Cesarean delivery is significantly increased in these women.10

For women with a history of placental abruption, the decision to conceive again arises several questions.5 Especially in the case of perinatal death, parents want to be reassured that this will not happen again. Although there are no interventions to prevent placental abruption, we must try to optimize the factors associated with placental abruption, for example by encouraging patients to quit any smoking and cocaine use. Furthermore we must strictly control hypertension in women trying to conceive or already pregnant. Since >50% of the recurrent placental abruptions occurred after 37 weeks of gestation, we would advise to plan an elective induction at this gestational age, nonetheless the neonatal outcome in the previous pregnancy.

In the Netherlands it is possible for pregnant women to deliver at home in case of a low risk pregnancy. Because women with a history of severe hemorrhage postpartum are at increased risk of recurrence, they have an indication to deliver in the hospital in a subsequent pregnancy. Although we found that women with severe postpartum hemorrhage after an early preterm delivery <32 weeks of gestation are not at increased risk of recurrence in a future pregnancy, this will probably not change antepartum management. Women with a previous preterm delivery <32 weeks of gestation will already be referred for hospital outpatient management because the increased risk of recurrence of premature delivery. The discussion on management in women with a history of postpartum hemorrhage continues and is limited by lack of definition and uniformity in research. Though our research on this subject uses the advised cut-off of 1000mL blood loss, information on shock, blood transfusion, hysterectomy, oliguria and coagulopathy are more informative parameters but often not available in databases and literature. Estimation of amount of blood loss always suffers from inaccuracies.11 As suggested by Kerr et al we must be more accurate in patients with (signs) of postpartum hemorrhage and we need different definitions for different purposes; one to decide when to start treatment, one for definition in research and one to use as quality indicator of care.12 Since postpartum hemorrhage is still a leading cause of maternal morbidity and mortality, women with suspicion of postpartum hemorrhage should be admitted to specialized centres and receive intensive monitoring and management in the current and subsequent pregnancies.


123

CHAPTER

8

Although this thesis answers some but not all questions regarding prenatal management in forestalled situations, our implications for clinical practice can be summarized as follows:- Actively search for (or exclude) vasa previa in women with one or more risk factors (Figure 1).- Withhold from premature Cesarean delivery in women with asymptomatic placenta previa but

schedule these women for elective Cesarean delivery between 37 and 38 weeks of gestation.

Figure 1. Flowchart on screening for vasa previa and management of vasa previa and placenta previa

Chapter 8

124

- Admission to the hospital and administration of corticosteroids in women with a symptomatic placenta previa and schedule Cesarean delivery at 35 or 36 weeks of gestation

- Schedule induction of labor at 37 weeks of gestation in women with a history of placental abruption and strictly control any form of hypertension.


125

CHAPTER

8

REFERENCES

1. Gianopoulos J, Carver T, Tomich PG, Karlman R, Gadwood K. Diagnosis of vasa previa with ultrasonography. Obstetrics & Gynecology. 1987 3/1987;69(3 Pt 2):488-91.

2. Oyelese KO, Turner M, Lees C, Campbell S. Vasa previa: an avoidable obstetric tragedy. [Review] [58 refs]. Obstetrical & gynecological survey. 1999 2/1999;54(2):138-45.



5. Rasmussen S, Irgens LM, Dalaker K. The effect on the likelihood of further pregnancy of placental abruption and the rate of its recurrence. Br J Obstet Gynaecol. 1997 Nov;104(11):1292-5.


7. Ioannou C, Wayne C. Diagnosis and management of vasa previa: a questionnaire survey. Ultrasound in Obstetrics & Gynecology. 2010 2/2010;35(2):205-9.

8. Lam CM, Wong SF, Chow KM, Ho LC. Women with placenta praevia and antepartum haemorrhage have a worse outcome than those who do not bleed before delivery. J Obstet Gynaecol. 2000 Jan;20(1):27-31.

9. Love CD, Fernando KJ, Sargent L, Hughes RG. Major placenta praevia should not preclude out-patient management. European journal of obstetrics, gynecology, and reproductive biology. 2004 Nov 10;117(1):24-9.

10. Johnston T.A.., S. P-B. Placenta Praevia, Placenta Praevia Accreta and Vasa Praevia: Diagnosis and Management. 2011 [cited; Available from:

11. Hancock A, Weeks AD, Lavender DT. Is accurate and reliable blood loss estimation the ‘crucial step’ in early detection of postpartum haemorrhage: an integrative review of the literature. BMC pregnancy and childbirth. 2015 Sep 28;15:230.

12. Kerr RS, Weeks AD. Postpartum haemorrhage: a single definition is no longer enough. BJOG : an international journal of obstetrics and gynaecology. 2016 Nov 24.

CHAPTER 9

Nederlandse samenvatting

Portfolio

Publicatielijst

Dankwoord

Curriculum Vitae


129

CHAPTER

9

NEDERLANDSE SAMENVATTING

AchtergrondHet uitdragen van een zwangerschap gevolgd door de bevalling van een gezond kind (of gezonde kinderen) is van diverse factoren afhankelijk. De ontwikkeling en groei van het ongeboren kind is cruciaal en kan onmogelijk plaatsvinden zonder placenta en navelstreng, welke zorgen voor een adequate uitwisseling van voedings- en afvalstoffen tussen moeder en kind. Een verminderde functie van de placenta is een relatief veel voorkomende complicatie van de zwangerschap en wordt gezien in ziektes als een zwangerschapshoge bloeddruk of zwangerschapsvergiftiging. Een suboptimale functie van de placenta kan leiden tot een groeivertraging van het kind, vroeggeboorte (< 37 weken zwangerschap) en daarmee ook een toename van het aantal ziekenhuisopnames van pasgeborenen. Maar, niet alleen een verminderde functie van de placenta kan leiden tot zwangerschapscomplicaties. Ook morfologische afwijkingen van de placenta zijn hiermee geassocieerd.

Dit proefschriftDit proefschrift bestaat uit twee delen en richt zich op een aantal morfologische afwijkingen van de placenta en de navelstreng en de daaraan gerelateerde complicaties voor moeder en kind. Het eerste deel richt zich op vasa previa ((vasa = bloedvat, previa = voorliggend), dit is een ernstige zwangerschapscomplicatie die gerelateerd aan een hoge mate van sterfte van de pasgeborene. Normaal gesproken vindt de navelstreng zijn oorsprong in het midden van de placenta. Bij vasa previa is dit niet het geval, en is de navelstreng enkel verbonden met de vliezen. Als deze vasa previa vóór de uitgang van de baarmoeder (de baarmoedermond) liggen, kunnen deze potentieel scheuren bij het breken van de vliezen. Omdat het bloed in deze vaten afkomstig is van het kind, en het kind slechts een totaal bloedvolume heeft van circa 300 – 400ml kan dit al snel fataal zijn voor het kind.

In het geval de placenta zélf voor de baarmoedermond ligt, is er sprake van een placenta previa. Deze complicatie gaat vaak gepaard met bloedverlies en ziekenhuisopnames tijdens de zwangerschap. Als de placenta tijdens de uitgerekende periode van de zwangerschap nog steeds voor de baarmoedermond ligt, kan er geen normale bevalling plaatsvinden en wordt een keizersnede verricht. Het tweede deel van dit proefschrift richt zich hierop, evenals op de placenta abruptio, ofwel het loslaten van de placenta vóór de geboorte van het kind. Dit is een potentieel ernstige complicatie welke nog regelmatig gevolgd wordt door het overlijden van het ongeboren kind. Als laatste onderdeel van het tweede deel van dit proefschrift wordt nog gekeken naar de kans op de herhaling van een vastzittende placenta na de bevalling, welke gepaard kan gaan met een (ernstige) toename van het bloedverlies bij de moeder (maternaal).

Chapter 9

130

In dit proefschrift over complicaties van de placenta en navelstreng wordt een antwoord gezocht op de volgende vragen:

- Hoe vaak komt vasa previa voor, wat zijn risicofactoren voor vasa previa en wat is de beste methode om vasa previa tijdens de zwangerschap vast te stellen?

- Welke vrouwen met een placenta previa hebben een verhoogd risico op een spoedkeizersnede?

- Wat is de kans op een vroegtijdige loslating van de placenta voor vrouwen die dit in een eerdere zwangerschap ook hebben gehad?

- Wat is de kans op een vastzittende placenta en ernstig maternaal bloedverlies na de bevalling voor vrouwen die dit in een eerdere zwangerschap ook hebben gehad?

Hieronder vatten we de belangrijkste bevindingen van elk hoofdstuk samen.

Vasa previaTegenwoordig ondergaan vrouwen, indien zij dit zelf ook wensen, meerdere echo’s tijdens hun zwangerschap. De mogelijkheid om vasa previa te onderzoeken met een echoscopisch onderzoek is in eerdere studies aangetoond. Echter, de betrouwbaarheid van deze diagnostiek is nooit onderzocht. In hoofdstuk twee hebben wij een uitgebreide literatuurstudie gedaan naar deze betrouwbaarheid. Wij constateerden in deze studie dat vasa previa betrouwbaar vastgesteld kan worden tijdens de zwangerschap, en wij hebben laten zien dat dit het beste kan gebeuren via een inwendige (transvaginale) echo ten tijde van de 20 weken echo (het structureel echoscopisch onderzoek ofwel SEO). In hoofdstuk drie hebben wij middels een tweede uitgebreide literatuurstudie gekeken naar het voorkomen van vasa previa en hebben wij risicofactoren voor deze complicatie onderzocht. Hieruit bleek dat één op de 1667 zwangerschappen gecompliceerd wordt door vasa previa en dat er meerdere risicofactoren zijn, te weten: een velamenteuze navelstrenginsertie, een meerlobbige placenta (placenta bilobata), een laagliggende placenta ten tijde van de 20 weken echo en een zwangerschap na een vruchtbaarheidsbehandeling (IVF of ICSI). In hoofdstuk vier hebben wij een richtlijn geschreven met betrekking tot vasa previa. Hierin wordt de diagnostiek, de risicofactoren en het beleid besproken om de zorg rondom vasa previa te optimaliseren. In dit hoofdstuk adviseren wij om bij zwangeren ten tijde van de 20 weken echo te beoordelen of er sprake is van een placenta previa, placenta bilobata of velamenteuze insertie en in die gevallen verder echoscopisch onderzoek te doen naar vasa previa met een inwendige echo. Als er vasa previa is vastgesteld bij een zwangere, adviseren wij in de richtlijn om een keizersnede te plannen vanaf 34 - 35 weken, met als voornaamste doel deze uit te voeren voordat de vliezen spontaan breken.


131

CHAPTER

9

Placenta previaHet huidige beleid bij een zwangere met een placenta previa ten tijde van de 20 weken echo, is om een echo te herhalen in het derde trimester van de zwangerschap. Vanwege de groei van de baarmoeder kan het zijn dat de placenta wordt opgetrokken en later in de zwangerschap dus niet meer voor de baarmoedermond ligt. In het geval de placenta nog wel voor baarmoedemrond ligt, zijn er diverse adviezen met betrekking tot het beste moment voor het verrichten van een geplande keizersnede. De meeste adviezen zijn gericht op een geplande keizersnede vlak voor de 37 weken zwangerschapsduur. In hoofdstuk vijf onderzochten wij in een cohort van Nederlandse zwangeren met placenta previa welke factoren geassocieerd zijn met een spoedkeizersnede (dat wil zeggen; het niet bereiken van de datum waarop de keizersnede oorspronkelijk gepland was). Uit dit onderzoek bleek dat een eerdere bevalling via een keizersnede, ziekenhuisopnames vanwege bloedverlies tijdens de zwangerschap en het ontvangen van bloedtransfusies ter compensatie van het bloedverlies, het risico op een spoedkeizersnede verhoogden. Tevens bleek uit dit onderzoek dat de vrouwen zonder klachten van bloedverlies of contracties gemiddeld bevielen bij een zwangerschapsduur van ongeveer 39 weken, en dat vroeggeboorte bij placenta previa dus mogelijk verminderd kan worden.

Placenta abruptioIn een groot nationaal cohort onderzochten wij in hoofdstuk zes hoe vaak en bij welke termijn een voortijdige loslating van de placenta voorkomt, evenals de kans dat dit nogmaals optreedt tijdens een volgende zwangerschap. Daarnaast onderzochten wij of het hebben van een hoge bloeddruk van invloed is op het optreden van een placenta abruptio. Wij vonden een incidentie van 0.22% onder 1,570,635 zwangere vrouwen. Deze incidentie was significant hoger onder vrouwen met een verhoogde bloeddruk, vergeleken met vrouwen met een normale bloeddruk. Voor een volgende zwangerschap bleek het risico op een placenta abruptio sterk verhoogd voor vrouwen met een eerdere abruptio (5.8%) vergeleken met vrouwen die dit niet eerder hadden meegemaakt (0.06%), met name voor vrouwen waarbij de eerdere abruptio plaatsvond voorbij de 37 weken zwangerschapsduur. Op basis van deze gegevens adviseren wij om vrouwen met een voorgeschiedenis van een placenta abruptio in een volgende zwangerschap in te leiden bij 37 weken zwangerschapsduur.

Retentio placenta en ernstig bloedverlies na de bevallingErnstig bloedverlies na de bevalling wordt gedefiniëerd als meer dan 1000mL en wordt regelmatig gezien bij een vastzittende placenta na de bevalling (retentio placenta). Vanuit eerdere studies is bekend dat vrouwen een verhoogd risico hebben op ruim bloedverlies als zij dit bij een vorige bevalling ook hebben gehad. In Nederland is een eerdere bevalling met ≥ 1000mL dan ook een reden om de volgende keer in het ziekenhuis te bevallen. In hoofdstuk

Chapter 9

132

zeven onderzochten wij de incidentie van ernstig bloedverlies en/of een vastzittende placenta in een groot nationaal cohort van zwangere vrouwen in Nederland en keken wij tevens of er verschillen waren in de kans op herhaling tussen vrouwen met ernstig bloedverlies met en zonder een retentio placenta,. Tevens keken wij naar de invloed van de zwangerschapsduur ten tijde van de eerste bevalling op de herhaalkans voor een volgende bevalling. In een groep van 359,737 vrouwen vonden wij een retentio placenta met en zonder ernstig bloedverlies in respectievelijk 2.1% en 5.4%. Het risico op deze complicaties in een volgende zwangerschap was voor de vrouwen met een voorgeschiedenis van ernstig bloedverlies bij de bevalling verhoogd (18%) vergeleken met vrouwen die dit niet hadden bij hun eerste bevalling (3.9%), behalve voor vrouwen die dit ernstige bloedverlies hadden bij een vroeggeboorte vóór 32 weken in hun eerste zwangerschap.

ConclusieHoewel dit proefschrift niet op alle vragen een antwoord kan geven, kunnen onze klinische implicaties als volgt worden samengevat:

- Zoek actief naar vasa previa bij vrouwen met één of meerdere risicofactoren- Plan een keizersnede bij vrouwen met een placenta previa niet voor 37 weken

zwangerschapsduur als zij geen klachten hebben- Neem zwangere vrouwen met een placenta previa en bloedverlies op in het ziekenhuis

ter observatie, overweeg corticosteroiden en plan een keizersnede rond 35 of 36 weken zwangerschapsduur gezien het hoge risico op een spoedkeizersnede in deze groep

- Laat vrouwen met een placenta abruptio in de voorgeschiedenis niet voorbij 37 weken zwangerschapsduur gaan bij een volgende zwangerschap, en behandel bij deze vrouwen een eventuele hypertensie.

Portfolio

133

CHAPTER

9

PHD PORTFOLIO

Courses- GeneralGraduate School, AMC:2016 Advanced topics in biostatistics2014 Scientific writing in English2014 Endnote course2014 Practical Biostatistics2013 Systematic Reviews2013 Basiscursus regelgeving en organisatie voor klinische onderzoekers (BROK)Erasmus MC, Rotterdam:2016 Advanced topics in decision making in medicine

Courses - Specific2013 Counseling prenatale screening 2013 Fetal medicine foundation: 1st trimester nuchal translucency2013 Fetal medicine foundation: midtrimester fetal anomalies scan

Seminars2013-2016 Weekly department seminars2014 Practicum foetale hartafwijkingen / fetal echocardiography, AMC2013-2015 Journal Club

Presentations2017 IGO Doelencongres, Presentation ‘screenen op vasa previa’, Rotterdam. 2016 SMFM The Pregnancy Meeting, Atlanta. Poster presentation ‘Incidence and

recurrence rate of severe postpartum hemorrhage and retained planceta: a longitudinal linked cohort study in the Netherlands’

2015 SMFM The Pregnancy Meeting, San Diego. Poster presentations ‘Incidence and recurrence rate of placental abruption: a longitudinal linked cohort study in the Netherlands’ and ‘Prediction of emergency delivery in women with placenta previa’

2014 ISUOG World Congress, Barcelona. Oral presentation ‘Accuracy of ultrasound in the diagnosis of vasa previa’

2014 Presentation ‘Dating of pregnancy’. Ultrasound course, Leiden University Medical Center

2014 Basis course on ultrasound in pregnancy ‘Placental and umbilical cord abnormalities’, AMC

Chapter 9

134

2014 WFE Annual meeting, presentation ‘Vasa previa’, Vredenburg Utrecht2013 WFE Annual meeting, presentation ‘Omphalocèle’, Vredenburg Utrecht

(Inter)national conferences2017 IGO – Doelencongres, Rotterdam2016 Society of maternal and fetal medicine (SMFM), Atlanta2015 Society of maternal and fetal medicine (SMFM), San Diego2014 International society of ultrasound in obstetrics and gynecology (ISUOG),

Barcelona

TeachingTutoring2016 - now Mentoring PhD student 2014, 2015 Teaching ultrasound and physical examination; medical students in their 2nd

year2013-2014 Student coaching/mentor scientific research project – medical bachelor student 2013-2014 Student coaching/mentor scientific research project – medical bachelor student

Publicatielijst

135

CHAPTER

9

PUBLICATIELIJST

Ruiter L, Mol BWJ, Pajkrt E. Authors’ reply re: Incidence of and risk factors for vasa praevia: a systematic review. British Journal of Obstetrics and Gynaecology. 2017 Jan;124(1):162-163.

Ruiter L, Eschbach SJ, Burgers M, Rengerink KO, Pampus van MG, Goes BY, Mol BW, Graaf IM, Pajkrt E. A prediction model for emergency caesarean section in women with placenta previa. American Journal of Perinatology. 2016 Dec;33(14):1407-1414.

Ruiter L, Kok N et al. Incidence of and risk indicators for vasa praevia: a systematic review. British Journal of Obstetrics and Gynaecology. 2016 Jul;123(8):1278-87.

Ruiter L, Ravelli et al. Incidence and recurrence rate of placental abruption: a longitudinal linked national cohort study in the Netherlands. American Journal of Obstetrics and Gynecology. 2015 Oct;213(4):573.e1-8.

Kok N, Ruiter L, Lindeboom R, de Groot C, Pajkrt E, Mol BW, Kazemier BM. Elective repeat cesarean delivery compared with trial of labor after a prior cesarean delivery: a propensity score analysis. European Journal of Obstetrics and Gynecol and Reproductive Biology. 2015 Dec;195:214-8.

Ruiter L, Kok N et al. Systematic review of accuracy of ultrasound in the diagnosis of vasa previa. Ultrasound in Obstetrics and Gynecology. 2015 May;45(5):516-22.

Kok N, Ruiter L, Hof M, Ravelli A, Mol BW, Pajkrt E, Kazemier B. Risk of maternal and neonatal complications in subsequent pregnancy after planned caesarean section in a first birth, compared with emergency caesarean section: a nationwide comparative cohort study. British Journal of Obstetrics and Gynaecology. 2014 Jan;121(2):216-23.

Dankwoord

137

CHAPTER

9

DANKWOORD

Ik ben blij dat ik de mogelijkheid heb gekregen om onderzoek te doen. Het maakt mij mede tot de persoon en dokter die ik op dit moment ben. Dit proefschrift is tot stand gekomen met hulp en aanmoediging van menig persoon. Ik wil in het bijzonder de volgende mensen bedanken.

Allereerst dr. Lips, beste Jos, zonder jouw aansporing was ik er misschien wel nooit aan begonnen. Dank voor je vertrouwen in mijn ‘klinische kunnen’ en je aanmoediging om mijn CV uit te breiden. Dank ook aan mijn oud-collega’s van het Kennemer Gasthuis, jullie zijn als een warm bad en maakten mijn keuze voor de Gynaecologie en Verloskunde definitief!

Eva, wat ben je een fantastisch persoon. Ik kan me ons eerste gesprek nog herinneren als de dag van gisteren; geen idee wat ik ervan moest vinden. Uiteindelijk bleek het goed en mocht ik bij je starten, ik ben je dankbaar voor alles wat je mij hebt bijgebracht op het gebied van echo. Jouw omgang met en vertrouwen naar patiënten toe zijn voor mij een groot voorbeeld. Ik hoop in de toekomst nog veel met je te mogen samenwerken!

Ben Willem, in onze fietstochten van en naar het AMC hebben we menig promotietraject uitgedacht. Veel dank voor je tijd, je begeleiding en je feedback, jij wist menig keer de vinger op de zere plek te leggen en zorgde ervoor dat ik het uiteindelijke doel voor ogen hield. Jij leerde me kritisch te zijn en met een andere oog naar onderzoeksresultaten te kijken.

Irene, jij stapte later in het traject in. Wij hadden op minder regelmatige basis contact, maar je was altijd behulpzaam en vooral vol vertrouwen in mij. Dank voor de mogelijkheid om onderzoek te doen naar de minder vaak voorkomende aandoeningen in ons vakgebied.

Bart en Larisa, van jullie heb ik het echoën écht geleerd. Dank voor jullie geduld, gezelligheid, kopjes koffie, stroopwafels en gedeelde lunches. Nienke, jij maakt me wegwijs in het onderzoek (en de congressen; Barcelona, sangria, strand en cocktails). Zullen we binnenkort weer eens gaan schaatsen op de Jaap Eden?

Anouk, Marie-Thérèse en Pascale, dank voor jullie gezelligheid en hulp op de polikliniek en tijdens de vele lunches! Ook dank aan alle andere dames op de polikliniek verloskunde en prenatale diagnostiek.

Stichting PRN, dank voor het gebruik van de data. Anita Ravelli, dank voor je hulp bij de analyses, menig uur hebben we eraan besteed maar met een prachtig resultaat.

Chapter 9

138

Bouchra, mijn onderzoeks- en echomaatje door dik en dun! Wat hebben we gepraat, gelachen, gehuild en veel meegemaakt de afgelopen jaren, zowel in het AMC, als daarbuiten (San Diego, roadtrip Californië, Atlanta, New York). Ik vind het onwijs jammer dat je straks niet aanwezig kunt zijn, maar kijk uit naar het moment dat je weer in Nederland bent! Ook veel dank aan mijn andere lieve kamergenoten van de afgelopen jaren; Maud (waardig echo-opvolger!), Merel, dank voor de gezelligheid op onze buitenlandse tripjes! Zullen we het nog eens overdoen…?

Angelique, dank voor je inspanningen tijdens de afrondende fase van mijn proefschrift, bij jou weet ik zeker dat de zaken in goede handen zijn.

Marchien, dank voor je steun en mogelijkheden die je geboden hebt om mijn proefschrift af te schrijven tijdens de start van mijn opleiding. Het viel niet altijd mee, maar het is maar mooi gelukt! Ik kijk uit naar het vervolg van mijn opleiding. Ook dank aan alle andere collega’s van het Flevoziekenhuis, ik heb al veel geleerd en ik hoop nog meer te leren in de komende jaren!

Toedels, dank voor jullie waardevolle vriendschap en vertrouwen. De frequentie van ons contact is soms wat minder, de waarde ervan zeker niet! Ik ben blij met vriendinnen zoals jullie en hoop dat onze vriendschap blijft bestaan tot in de eeuwigheid! Louise, heerlijk om onze onderzoeksperiode in het AMC te hebben kunnen delen, wekelijks sparren onder het genot van koffie heeft ons allebei veel opgeleverd denk ik! Heel veel succes met de afronding van je eigen promotie en tot snel in het AMC! Lieve meiden, zullen we gauw weer een vakantie boeken?

Al mijn andere vrienden en vriendinnen, dank voor jullie luisterend oor als ik het eens nodig gehad, geduld als ik ergens niet bij kon zijn maar vooral dank voor jullie gezelligheid en vertrouwen! Ellen; altijd een goed (en opbeurend!) woord heb je voor me klaar, ik kijk uit naar al onze uitstapjes in de toekomst (met en zonder de kleine meisjes erbij).

Mijn paranifmen; Chantal, vanaf het moment dat je bij ons op de kamer kwam, was het dikke mik. Ruim anderhalf jaar zaten we tegenover elkaar, één blik was voldoende om van elkaar te weten hoe het ging. Dank voor alle cappuccino's, gezelligheid, slappe lach, lieve woorden en steun. Ik ben blij dat jij straks naast mij staat en kijk uit naar het moment dat ik je weer als collega mag verwelkomen in het Flevo!En Marinka, ik kan wel zeggen; door dik en dun! Wat hebben wij al veel meegemaakt, hoeveel kilometer zouden we gezamenlijk al hebben afgelegd? Ik vind het fantastisch hoe jij – als niet-medicus – altijd geïnteresseerd bent in mijn werk en alles wil weten van mijn onderzoek. Ik ben blij dat jij straks naast mij staat!

Dankwoord

139

CHAPTER

9

Lieve papa en mama, (schoon)broers en –zussen, Betty; dit is nu waar ik het al die tijd over had, en het is klaar! Dank voor jullie support, lieve woorden en gezelligheid!

Lieve Thijs, dank voor je eeuwige geduld en vertrouwen. Al vanaf het eerste begin stond je me bij en je hebt het nooit opgegeven. Ook zonder medische kennis wist je me te helpen als ik vastliep. Dank ook voor alle fijne en ontspannen momenten die je me bood, voornamelijk tijdens de laatste fase van mijn promotie had ik die soms echt even nodig. Je bent werkelijk de allerliefste en ik verheug me op de rest van ons leven samen!Liefste Nina, vanaf nu zijn mijn vrije dagen écht vrij en helemaal voor jou. Laten we nooit ophouden met samen spelen, lezen, puzzelen, fietsen, zwemmen, knuffelen en al van dat nog meer!

Curriculum Vitae

141

CHAPTER

9

CURRICULUM VITAE

Laura Ruiter werd op 25 juni 1986 geboren in Naarden als jongste van twee kinderen en groeide op in Huizen en Almere. In 2004 behaalde zij haar VWO diploma aan het Oostvaarders College te Almere. Na uitgeloot te zijn voor Geneeskunde deed zij kortdurend de studie Biomedische Wetenschappen aan de Universiteit van Amsterdam, tot zij in september 2005 alsnog kon starten met de studie Geneeskunde aan de Universiteit van Amsterdam. Na het doctoraalexamen volgde zij haar coschappen, haar keuze-co schap was een combinatie van kindergeneeskunde en verloskunde en deze deed zij in India. In 2012 behaalde zij haar artsexamen en vanaf maart 2012 werkte zij een jaar als ANIOS gynaecologie in het Kennemer Gasthuis in Haarlem. Onder leiding van Prof. dr. E. Pajkrt, prof. dr. B.W.J. Mol en Dr. I.M. de Graaf deed zij van maart 2013 tot april 2016 promotie onderzoek in het Academisch Medisch Centrum, waarvan de resultaten zijn beschreven in dit proefschrift. Gedurende deze periode werkte zij tevens als arts-echoscopist op de poli Verloskunde en Prenatale Diagnostiek van het AMC te Amsterdam. In september 2016 startte zij met het eerste deel van de opleiding tot gynaecoloog in het Flevoziekenhuis te Almere.

UITNODIGING

voor het bijwonen van de openbare verdediging van het

proefschri�



op woensdag 22 november 2017 om 12.00 uur

in de Agnietenkapel van de Universiteit van Amsterdam Oudezijds Voorburgwal 231

te Amsterdam.

Na afl oop bent u van harte welkom bij de recep� e

ter plaatse.

Laura RuiterDodaarsweg 33

3897 LP Zeewolde06-49320999

[email protected]

Paranimfen:Chantal Diedrich

06-48506833

Marinka Bierma-Enthoven06-15662653

Vasa previa and placenta associated complications Diagnosis and risk assessment

Laura Ruiter

Vasa previa and placenta associated com

plications - diagnosis and risk assessment L

aura Ruiter

Ruiter_cover.indd Alle pagina's 6-10-2017 11:59:31

UvA-DARE (Digital Academic Repository) Vasa previa and ... · Chapter 3 Incidence of and risk...

Documents

Transcript of UvA-DARE (Digital Academic Repository) Vasa previa and ... · Chapter 3 Incidence of and risk...