6. Jahrgang 2009 // Nummer 2 // ISSN 1810-2107

Journal für

Reproduktionsmedizinund Endokrinologie

No.2 2009

– Journal of Reproductive Medicine and Endocrinology –

www.kup.at/repromedizin

Online-Datenbank mit Autoren- und Stichwortsuche

Krause & Pachernegg GmbH, Verlag für Medizin und Wirtschaft, A-3003 Gablitz

Offizielles Organ: AGRBM, BRZ, DIR, DVR, DGA, DGGEF, DGRM, EFA, OEGRM, SRBM/DGE

Indexed in EMBASE/Excerpta Medica

Andrologie • Embryologie & Biologie • Endokrinologie • Ethik & Recht • GenetikGynäkologie • Kontrazeption • Psychosomatik • Reproduktionsmedizin • Urologie

Member of the

Oral Contraceptives and Venous Thromboembolism -

Consensus Opinion from an International Workshop held

in Berlin, Germany in December 2009

Reid RL, Westhoff C, Mansour D, de Vries C

Verhaeghe J, Boschitsch E, Gompel A, Birkhäuser M

Krepelka P, Dulicek P, Iversen OE, Khamoshina M

Dezman LV, Fruzzetti F, Szarewski A, Wilken-Jensen C

Seidman D, Kaaja R, Shapiro S

J. Reproduktionsmed. Endokrinol 2010; 7 (Sonderheft

1), 67-72

Unser Laborteam: Inkubatoren von LabotectQualität - Made in Germany

Benchtop InkubatorLabo C-Top

CO -Inkubatoren2

C16 C60

www.labotect.comsales@labotect.com+49 551 / 50 50 125

Fragen Sie VOR Ihrer Kaufentscheidung nach den Zertifikaten!

Vertiefende Informationen zum Thema Zertifizierung senden wir Ihnen gerne auf Anfrage zu.

C200

Oral Contraceptives and Venous Thromboembolism

J Reproduktionsmed Endokrinol 2010; 7 (Special Issue 1) 67

Oral Contraceptives and Venous ThromboembolismConsensus Opinion from an International Workshop

held in Berlin, Germany in December 2009 *

R. L. Reid, C. Westhoff, D. Mansour, C. de Vries, J. Verhaeghe, E. Boschitsch, A. Gompel, M. Birkhäuser, P. Krepelka, P. Dulicek,O. E. Iversen, M. Khamoshina, L. V. Dezman, F. Fruzzetti, A. Szarewski, C. Wilken-Jensen, D. Seidman, R. Kaaja, S. Shapiro

* Reprint from: J Fam Plann Reprod Health Care 2010; 36 (3): 117–22. Reprinted with permission.Correspondence: Professor Robert L Reid, Department of Obstetrics and Gynecology, Faculty of Health Sciences, Queen’s University, Kingston, Ontario, Canada ON K7L 4V7;e-mail: robert.reid@queensu.ca

Robert L. Reid – Department of Obstetrics and Gynecology,Faculty of Health Sciences, Queen’s University, Kingston,Ontario, Canada

Carolyn Westhoff – Department of Obstetrics and Gyneco-logy, College of Physicians and Surgeons, Columbia Univer-sity, New York, NY, USA

Diana Mansour – NHS Newcastle and North Tyneside,Community Health, New Croft Centre, Newcastle uponTyne, UK

Corrine de Vries – Department of Pharmacy and Pharmaco-logy, University of Bath, Bath, UK

Johan Verhaeghe – Department of Obstetrics and Gyneco-logy, Katholieke Universiteit Leuven, Leuven, Belgium

Ewald Boschitsch – Ambulatorium KLIMAX, Wien, Austria

Anne Gompel – Service de Gynécologie, Université ParisDescartes, Paris, France

Martin Birkhäuser – Department of Obstetrics and Gynae-cology, University of Bern, Switzerland

Petr Krepelka – Institute for Care of Mother and Child,Podolske nabrezi, Prague, Czech Republic

Petr Dulicek – Division of Clinical Haematology, UniversityHospital, Hradec Králové, Sokolská, Czech Republic

Ole-Erik Iversen – Institutt for Klinisk Medisin, Seksjon forGynekologi og Obstetrikk, Kvinneklinikken, HaukelandUniversitetssykehus, Bergen, Norway

Marina Khamoshina – Department of Obstetrics and Gyne-cology, RUDN/People Friendship University of Russia, Mos-cow, Russia

Lucija Vrabic Dezman – Gynaecological Clinic, HealthCentre Kranj, Gosposvetska, Kranj, Slovenia

Franca Fruzzetti – Department of Obstetrics and Gyneco-logy, Ospedale S Chiara, Pisa, Italy

Anne Szarewski – Cancer Research UK Centre for Epidemi-ology, Mathematics and Statistics, Wolfson Institute of Pre-ventive Medicine, London, UK

Charlotte Wilken-Jensen – Region Sjælland, Gynæko-logisk-Obstetrisk Afdeling, Roskilde Sygehus, Køgevej,Roskilde, Denmark

Daniel Seidman – Department of Obstetrics and Gyneco-logy, Sheba Medical Center, Tel-Hashomer, Sackler Schoolof Medicine, Tel-Aviv University, Tel-Aviv, Israel

Risto Kaaja – Turku University, Satakunta Central Hospital,Pori, Finland

Samuel Shapiro – Department of Public Health and FamilyMedicine, University of Cape Town, Cape Town, South Africa

Workshop Participants/Consensus Opinion Contributors

Background and Purpose

of the Workshop

Concern about the venous thrombo-embolism (VTE) risk of new hormonalcontraceptive options shortly after theirentry into the market has triggered anumber of ‘pill scares’, each of which re-sulted in panic stopping of the formula-tions in question and a spike in un-planned pregnancies, yet with no subse-quent reduction in VTE rates amongwomen of reproductive age.

Perhaps the best example of a recent pillscare that resulted in enormous harm

from a public health perspective was the‘third-generation pill scare’ that oc-curred in many countries in Europe andaround the world in 1995. At that timethe new third-generation pills were pro-moted as being less androgenic and aspossibly having fewer adverse effects oncardiovascular and metabolic param-eters and thereby potentially being saferthan existing pills.

Shortly after the introduction of thesethird-generation pills, reports of a pos-sible increased risk of VTE began to ap-pear. Such reports brought the progesto-gen component of these pills under scru-

tiny and inevitably a phenomenonknown as “stimulated reporting” oc-curred. Physicians with patients on thesenew pills were more likely to send theirpatients for assessment of any leg pain orswelling and more likely to report anyVTE episodes to regulatory authoritiesbecause of heightened awareness of thepossible risk. What followed was an in-ternational pill scare when regulatoryauthorities in a number of countries is-sued alerts about the possible increasedrisk of VTE with thirdgeneration pills.Panic stopping of pills by millions ofwomen resulted in an abrupt and alarm-ing rise in unplanned pregnancy as evi-

For personal use only. Not to be reproduced without permission of Krause & Pachernegg GmbH.

68 J Reproduktionsmed Endokrinol 2010; 7 (Special Issue 1)

Oral Contraceptives and Venous Thromboembolism

denced by sudden increases in deliveriesand abortions [1–3], each with their at-tendant increased risks of VTE.

The history of this unfortunate episode iswell documented. After class action law-suits and lengthy trials, at which expertsin epidemiology debated the findings,there emerged an awareness that studiesexamining VTE risks, in particular, re-quired an understanding of the epidemi-ology of VTE and the possible biasesthat might systematically result in find-ings which diverged from the “truth” [4,5]. The precise effects of different hor-monal contraceptives on the haemostaticsystem continue to be studied and de-bated [6, 7] but because compelling datafor an increased risk have not been dem-onstrated the lawsuits were ultimatelythrown out [8] and third-generation pillsremain on the market and are widely pre-scribed today [9].

In 2009, anecdotal reports of VTE epi-sodes in women using an ethinylestradi-ol/drospirenone (EE/DRSP)-based oralcontraceptive (Yasmin®) were followedby publications with conflicting findingsabout the risk of VTE with this product.This led the manufacturer to invite ex-perts in gynaecology, reproductive en-docrinology, haematology and epidemi-ology to a workshop to critically reviewthese recent publications to understandthe reasons for divergent results and toproduce a consensus statement about thelikely comparative risk of this new for-mulation and other marketed products.Those contributing to the workshopcritically appraised all relevant papers ina consistent manner, applying a stan-dardised critical appraisal tool to ensureequity and consistency between apprais-als of studies [10]. This document is theresulting consensus opinion, drafted byall the experts invited to the workshopand unedited by the manufacturer’s em-ployees.

If certain combined oral contraceptives(COCs) carry a higher VTE risk thanothers, this needs to be clear so womencan be advised accordingly and well-in-formed treatment decisions can be made.At the same time, scares founded onflawed study results that become drivenby media and legal interests do not servethe public good. It is essential that thelevel of risk is determined through abestevidence evaluation of available

data and that balanced and informed de-cisions are made by considering bothbenefits and risks of contraceptive for-mulations.

Oral Contraceptives: Bal-

ancing Benefits and Risks

Modern COCs afford not only excellentcontraception but also a variety of non-contraceptive benefits ranging fromregulation and reduction of both men-strual bleeding and dysmenorrhoea totreatment of premenstrual syndrome,menstrual migraines, acne and hirsut-ism. Long-term benefits include reducedrates of endometrial and ovarian cancer[11].

Serious side effects are rare occurrencesin OC users. Modern COCs are well tol-erated and adherence to prescribed regi-mens is generally excellent thus allow-ing women to postpone their pregnan-cies. By avoiding unplanned pregnan-cies, women reduce their exposure to se-rious vascular risks and flare-ups of sys-temic diseases that may occur duringpregnancy.

Venous Thromboembo-

lism (VTE)

VTE remains a rare but potentially seri-ous complication of COC use. This con-dition typically involves thrombosis inthe deep veins of the legs or pelvis andthe potential for pulmonary embolism,which has potentially fatal consequences.Known risk factors for VTE include ad-vancing age [12], antiphospholipid anti-bodies and hereditary thrombophilia [13,14], cigarette smoking [15], surgicalprocedures [16] trauma, and immobility(such as that associated with travel orhospitalisation) [17], obesity [18], andpregnancy and the peripartum period[19, 20].

Quoted rates of VTE appear to have in-creased in the general population in thepast decade from 1/10,000 woman-yearsfor non-COC users to 4–5/10,000 womanyears during the reproductive years [21].This may in part reflect a true increase inVTE rates due to demographic trendssuch as increasing obesity but also re-flects greater physician awareness, in-creased referral to hospitals and im-provements in diagnostic precision (e. g.Doppler ultrasound).

Traditionally hormonal contraceptiveswere thought to double or triple the rateof VTE above the background rate. Whilea decade ago common teaching was thatnon-pill users had VTE rates of 1/10,000woman-years and COC users had ratesof 2–3/10,000 woman-years, moderndata obtained through active surveil-lance have provided new insights intothe true risks. Contemporary evidencesuggests that VTE rates in non-pregnant,non-pill users of reproductive age arehigher than previously thought – about4–5/10,000 woman-years in non-COCusers [21] with an approximate doublingof risk into the range of 9–10/10,000woman-years for COC users [22].

To keep the risks of VTE for COC usersin perspective it is important to remem-ber that the risk of a VTE in pregnancy isas high as 29/10,000 woman-years [22,23] and in the immediate postpartumperiod may reach 300–400/10,000woman-years [19, 20]. As one of themost widely used and effective contra-ceptive methods, the pill reduces rates ofunplanned pregnancies and actually de-creases the overall rate of VTE in thepopulation in comparison to populationswithout access to effective contraception[24].

Past research has shown that older,higher estrogen dose pills carried aslightly greater risk than currently mar-keted OCs, most of which contain lessthan 50 µg EE. Sub-50 µg pills have alower risk of VTE than pills with 50 µgEE or more [25]. While, in theory,greater reductions in the dosage of estro-gen might further decrease the risk ofVTE, this benefit has not been clearlyestablished. A non-significant decreasein VTE was noted in the EURAS Studywhen comparing 30 µg to 20 µg EE-con-taining pills [22]. Lidegaard also re-ported that “a reduction in estrogen dosefrom 30–40 µg to 20 µg for OCs contain-ing desogestrel or gestodene reduced therisk of venous thromboembolism by18 % (7–27 %)” [26]. However, thevalidity of the VTE diagnoses in theDanish registry has been questioned[27].

Pills with 20 µg EE or less have the po-tential to cause more breakthrough (un-scheduled) bleeding and this may be adeterrent to consistent use for somewomen [28].

Oral Contraceptives and Venous Thromboembolism

J Reproduktionsmed Endokrinol 2010; 7 (Special Issue 1) 69

Progestogen-only contraceptive meth-ods have not been shown to increase therate of VTE [29]. Transdermal deliveryof estradiol in hormone therapy formenopausal women may be associatedwith a reduced risk of VTE [30] butavailable data do not suggest a benefit interms of VTE risk when non-oral routes(i. e. transdermal and intravaginal) areused for delivery of combined hormonalcontraception using EE [31].

Innovations in hormonal contraceptionin recent years have largely involved theuse of new progestogens – not becausethere was a belief that these progesto-gens might influence VTE rates, butrather to harness the additional anti-an-drogenic benefits of some of these newproducts.

Assessing VTE Risks of

New Oral Contraceptive

Products

Surrogate Markers for VTE RiskWorkshop participants considered theconcept of “estrogenicity” of OCs asmeasured by the differential impact ofdifferent estrogen and progestogen com-binations on hepatic protein synthesis(primarily sex hormone binding globulin[SHBG] but also ceruloplasmin and cor-ticosteroid binding globulin [CBG]) andcertain coagulation parameters (such asactivated protein C [APC resistance andtotal and free protein S levels]).

In postmenopausal women differentroutes of estrogen administration appearto be associated with different effects onhepatic protein synthesis and possiblydifferent risks for VTE [30–33]. The ap-parent differential effects of these routesof estrogen delivery led to the conceptthat “estrogenicity” (as determined byhepatic production of SHBG in responseto treatment) might predict risk for VTE.

Since the controversy about the VTErisk with third-generation OCs surfacedin 1995, numerous investigators haveevaluated differences in hepatic proteinsynthesis with different OC formula-tions to try to shed light on purported dif-ferences in VTE risk. Differential effectson hepatic proteins and APC resistance(which parallels the changes in SHBG)have been documented for different OCformulations, lending support to the con-cept of differential “estrogenicity”, with

an implied association between third-generation pills (which affect these pa-rameters more than pills containinglevonorgestrel [LNG]) and risk for VTE.While these findings have been used totry to explain why certain OC formula-tions might carry greater risks for VTE[34] the clinical evidence to confirmsuch speculation is presently lacking. Tobe a valid surrogate marker for diseasethe marker must be validated in at leastone prospective trial using both the sur-rogate marker and the true outcome [35]and this has never been done [36]. Atpresent there are no validated surrogatemarkers for VTE risk associated withhormonal contraception.

Premarketing Indicators of VTERiskVTE remains one of the few serious ad-verse effects of contemporary hormonalcontraception. Because VTE is so rare,valid information about the risks associ-ated with a given pill cannot be ad-equately gleaned from premarketing re-search, which typically involves 1000–3000 women. As a result, any informa-tion on serious but rare side effects likeVTE must come from postmarketingsurveillance.

Postmarketing Surveillance:Voluntary Reporting of AdverseReactionsMany countries employ a voluntary sur-veillance system in which health careproviders are asked to report unexpectedand/or severe adverse reactions of drugsto regulatory authorities once theseproducts enter the marketplace. Al-though such a system can alert regula-tory authorities to rare unexpected sideeffects of a medication, the system ishighly dependent on busy clinicians torecognise a potential adverse drug reac-tion and then to locate and complete thenecessary forms when adverse events areencountered. Many times events go un-reported and at times the same event maybe reported by more than one physician.Media coverage of a particular concernis a well known cause for “stimulated re-porting” and self-referral among womenusing the product in question.

As a result, such a system is unableto provide valid information on the num-ber of cases (numerators) for particularproducts. Since denominators (the num-ber of women using a particular product)

and the distribution of risk factorsamong the entire population of womenprescribed a specific type of OC areunknown, voluntary case reports cannotprovide valid information on compara-tive risks for rare but serious side effects.

Observational Studies andClinical TrialsFor both observational studies and clini-cal trials the importance of risk factorassessment in populations under studycannot be overestimated. For example, ifa product is perceived as having specificnon-contraceptive benefits for obesewomen (DRSP may limit cyclic monthlyweight gain and reduces acne and facialhair growth – both common worries foroverweight women) then it may be pref-erentially prescribed to this population,which as a whole has a slightly higherrisk of VTE. Such was the case withDRSP-containing pills as demonstratedin the non-interventional European Ac-tive Surveillance Study. In this large co-hort study, health care providers wereallowed to choose the OC they thoughtmost appropriate in individual circum-stances. Data analysis at the conclusionof the study showed that DRSP-contain-ing pills were preferentially prescribedto obese women [22].

One approach to obtaining informationon rare events after marketing of a newproduct is to examine large databases tolook for usage of a specific product andto establish linkages to complicationslike VTE. These techniques may be use-ful if the original data set collected infor-mation on age, body mass index (BMI),duration of use, family history of VTE,and so on, which are important indepen-dent risk factors for VTE. In the absenceof quality data on such variables, it isimpossible to exclude the possibility thata risk apparently associated with a spe-cific OC is due to characteristics of thewomen to whom it was prescribed ratherthan a result of exposure to specific con-stituents of that pill. Databases, espe-cially those developed for administrativepurposes, may lack adequate validationof discharge diagnoses. Recent researchhas shown that conditions entered asVTE in an administrative database wereoften something else when testing wascompleted [27].

A valid comparison of VTE risk withdifferent OCs must consider the fact that

70 J Reproduktionsmed Endokrinol 2010; 7 (Special Issue 1)

Oral Contraceptives and Venous Thromboembolism

the risk for VTE is greatest in the firstmonths of use and then falls closer to thebackground risk but remains elevated[22, 37–39]. This same effect was notedin the largest randomised clinical trial toexamine VTE in menopausal women re-ceiving hormone therapy at the time ofmenopause [40]. Comparisons of theVTE risks between different OCs mustaccount for recency of initiation ofCOCs and ensure that new users of onepill are truly being compared to newusers of the other pill.

The type of research that is most likely toprovide legitimate information on theserare risks involves a prospective ap-proach with all subjects being new usersof an OC. Careful follow-up by activesurveillance (regular patient contact)and validation of all suspected cases(through examination of the medicalrecords) is critical.

At the time of introduction of a new EE/DRSP-containing COC into Europe, themanufacturer – with the approval ofEuropean regulatory authorities – initi-ated a study to compare the VTE risk ofthis new product with other marketedCOCs in a “real life” situation. Simulta-neously, 2 database studies were carriedout as well as a post hoc analysis of acase-control study set up with a differentprimary study aim. We provide an over-view in the following section, in whicheach of the studies is described togetherwith an overview of the study results,strengths, limitations and our resultinginterpretation of the key findings.

Recently Published Studies

European Active SurveillanceStudy (EURAS)The European Active Surveillance Study(EURAS) was conducted with the ap-proval of European regulatory authori-ties and with oversight by an indepen-dent advisory board. This prospective,non-interventional, active surveillancestudy followed 59,674 new users ofdifferent OCs for a total of 142,475woman-years of follow-up.

Active follow-up resulted in a “loss-to-follow-up” of only 2.4 %. All diagnoseswere validated by review of medicalrecords. The study demonstrated thatDRSP-containing OCs were prescribedmore often to heavier women and that

despite this there was no difference inVTE rates between this new DRSP-con-taining OC and other marketed OCs[22].

Ingenix™ Claims DatabaseA second large prospective study on theIngenix™ claims database in the USAused a propensity (risk factor) scoringsystem to match new users of differentpills according to baseline VTE risk [41,42]. This study matched new OC pre-scriptions (Yasmin vs other pills in a 2:1ratio) and validated the diagnoses byexamination of medical records of allpotential VTE episodes. No differencesin VTE rates were found betweenDRSP-containing OCs and other mar-keted pills.

Controversy over Recent Publi-cationsPublicity following reports of VTE epi-sodes in users of Yasmin has once againignited the debate about whether somefeature of the new progestogen, drospi-renone, may contribute to an increasedrisk of VTE compared to other pills.Two reports published in the BritishMedical Journal in 2009 appear, at firstglance, to support this fear. Criticalanalysis of the 2 studies responsible forthis adverse publicity [26, 43] however,suggests that the conclusions could wellhave resulted from methodologicalflaws and/or misinterpretation of find-ings [7, 44].

Dutch Mega StudyThe Dutch Mega Study [43] was a studydesigned to evaluate environmental andgenetic factors influencing VTE. Theauthors performed a substudy using acase-control approach to examine the ef-fects of different OCs on VTE. Caseswere identified from women with VTEattending an anticoagulation clinic. Con-trols were found in an unusual way, withmany being partners of men seen inclinic and the remainder being foundthrough random digit dialling. Durationof use information was not available onall women so it is not clear that all werenew users, and women were not matchedaccording to BMI. The authors showedhazard ratios for various OCs with wideand overlapping 95 % confidence inter-vals (CIs) indicating no statistically sig-nificant difference in the VTE risk be-tween DRSP-containing pills and otherpills. Despite this they concluded that

second-generation OCs were safer as faras VTE was concerned. In recent revi-sions to the Yasmin product label in theUSA, the Food and Drug Administration(FDA) has concluded: “The number ofYasmin cases (in the Dutch Mega Study)was very small (1.2 % of all cases) mak-ing the risk estimates unreliable.” [45].

Danish National Cohort StudyThe other study reporting increased ratesof VTE with DRSP-containing andthird-generation OCs was a large DanishNational Cohort Study [26]. Prescrip-tions of Danish women and subsequenthospital diagnostic codes for VTE wereretrospectively examined between 1995and 2005 in a number of interrelateddatabases maintained by the Danishnational health service. This comprehen-sive data set included 3.3 millionwoman-years of OC use. The investiga-tors identified 2045 OC-associatedVTEs during this time period. In keepingwith prior reports they found that the riskof VTE was greatest shortly after initiat-ing COC use and was lowest with pillscontaining the lowest estrogen dosages.In contrast to the prospective studiesreported above they found that DRSP-containing OCs and third-generationOCs (i. e. those containing desogestrelor gestodene) carried an increased riskof VTE compared to LNG-containingOCs.

Several significant methodological weak-nesses have been identified since thepublication of that report [44]. First, de-tailed information on confounders suchas obesity, surgery or trauma was notavailable for the analysis. This is impor-tant as in the Danish population there hasbeen close to a three-fold increase inobesity in the past decade, and evidencefrom other research suggests that obesewomen are more likely to be prescribedDRSP-containing OCs.

In addition, the investigators had no in-formation on OC use before 1995.Lidegaard et al. classified all LNG usersin 1995 as “short-term” users. In re-sponse to the criticism that this mighthave unfairly compared long-term LNGusers to short-term users of DRSP-con-taining pills Lidegaard responded that “itwas very unlikely that this small mis-classification of short-term users startingtheir short-term use in the beginning of1995 would have had a substantial influ-

Oral Contraceptives and Venous Thromboembolism

J Reproduktionsmed Endokrinol 2010; 7 (Special Issue 1) 71

ence on our estimates” [46]. Others haveestimated that as many as 60 % of thoseLNG users classified as having “short-term” exposure in 1995 were misclassi-fied, and that these individuals shouldhave been classified as “long-term” us-ers [47]. Such a misclassification couldhave significantly biased results infavour of LNG-containing COCs. Sup-porting the premise that the duration oftime since initiation of COC use wasmisclassified for many LNG users wasthe unexpected finding that all OCs withthe exception of those containing LNGshowed the anticipated first-year eleva-tion of VTE risk.

A recent publication from Severinsenand colleagues in Denmark suggests thatthe incidence of VTE could not reliablybe assessed in the Danish registry [27].These authors examined the same Dan-ish registry used by Lidegaard’s group,and after reviewing the medical recordsof 1135 patients with a registry diagnosisof VTE they could confirm the diagnosisin only 31 % of cases referred from emer-gency rooms and in only 71 % of womenadmitted to the ward for diagnostic test-ing. The Danish database was designed asan administrative database rather than asa medical research database. The diag-nostic codes were primarily intended tocapture costs related to hospitalisation.Many physicians were incorrectly enter-ing a diagnostic code of VTE rather thanthe intended code of “admitted for evalu-ation of possible VTE”.

The increased rate ratio of VTE withDRSP-containing OCs compared toLNG-containing OCs was 1.64 (95 %-CI: 1.27–2.10). Considering the potentialmethodological issues identified above,and with rate ratios that in epidemiologi-cal terms are very small (relative risks of2 or less), it is extremely difficult to ex-clude bias or residual confounding as theexplanation for the findings [7, 44, 48, 49].

The recent serious allegations about thevalidity of the VTE diagnostic codes inthe Danish database throw into doubtany conclusions about VTE risk of dif-ferent pills based on this source. Thenew US FDA-endorsed product label-ling for Yasmin addresses the Lidegaardstudy and concludes: “The risk estimatesmay not be reliable because the analysismay include women of varying risk lev-els.“ [45].

Conclusions

The highest quality scientific studiesevaluating the risk of VTE in womenof childbearing age show a risk of ap-proximately 4–5/10,000 woman-years inthose who do not use OCs. In the ab-sence of reliable contraception, womenof reproductive age face risks of VTEassociated with pregnancy and the im-mediate postpartum period that reach300–400/10,000 woman-years. Prospec-tive observational studies have shownthat all currently marketed COCs in-crease the risk of VTE to the range of9–10/10,000 woman-years of use andthat this risk is highest in the first monthsof use with a fall towards baseline riskthereafter. Modern OCs offer excellentcontraceptive efficacy and good adher-ence due to their many non-contracep-tive benefits.

Although VTE related to OC use is a rareevent, future research should strive toreduce the VTE rate further. Clinicalprediction models for incident VTEshould be developed on the basis ofclinical parameters (age, BMI, smoking,family history, etc.) or clinical param-eters plus screening for genetic thrombo-philia defects in women with a positivefamily history. Each of the thrombo-philia tests should be subjected to a cost-effectiveness analysis. In addition, thevalue of the proposed biochemical mark-ers of VTE in OC users (APC resistance,SHBG, etc.) must be assessed in largeclinical studies, and the assays need to bestandardised by international consensus.Finally, the VTE risk associated withfurther therapeutic innovations (e.g. es-tradiol-based OCs, non-oral formula-tions, etc.) should be evaluated.

Individual adverse events (such as pul-monary embolism or stroke) in womenon combined hormonal contraception,though rare, are extremely unfortunateand must always be taken seriously. Pub-lic health is best served when medicalmanagement is based on the highestlevel of scientific evidence and by care-fully considering whether any increasesin risk outweigh the benefits of treatment.

For most women the very low absoluterisks of serious sequelae are outweighedby the well-established benefits of hor-monal contraception. Research hasshown that individualised risk assess-

ment remains an important tool to iden-tify women at increased risk for VTEwho might be better advised to use alter-native forms of contraception. For mosthealthy women of reproductive age thebenefits of COCs will outweigh the risks.

Statements on Funding

and Competing Interests

Funding: The workshop from which thisconsensus opinion arose was convenedby Bayer Schering. None of the invitedworkshop participants/contributing au-thors has received any funding in respectof preparation of this consensus opinion,which has been produced/published com-pletely independently of Bayer Schering.

Competing interests: The majority ofthe workshop participants/contributingauthors will, at one time or another, havereceived sponsorship and/or fundingfrom one or more manufacturers of con-traceptive drugs and products.

References:

1. Mills A. Combined oral contraception and the risk of venousthromboembolism. Hum Reprod 1997; 12: 2595–8.2. Goodyear-Smith A, Arroll B. Termination of pregnancy fol-lowing panic-stopping of oral contraceptives. Contraception2002; 66: 163–7.3. Skjeldestad FE. Increased number of induced abortions inNorway after media coverage of adverse vascular events fromthe use of third-generation oral contraceptives. Contraception1997; 55: 11–4.4. Spitzer WO. The aftermath of a pill scare: regression to re-assurance. Hum Reprod Update 1999; 5: 736–45.5. Lewis MA, MacRae KD, Kuhl-Habichl D, Bruppacher R,Heinemann LA, Spitzer WO. The differential risk of oral con-traceptives: the impact of full exposure history. Hum Reprod1999; 14: 1493–9.6. Hennessy S, Berlin JA, Kinman JL, Margolis DJ, MarcusSM, Strom BL. Risk of venous thromboembolism from oralcontraceptives containing gestodene and desogestrel versuslevonorgestrel: a meta-analysis and formal sensitivity analy-sis. Contraception 2001; 64: 125–33.7. Dinger J. Oral contraceptives and venous thromboembo-lism: old questions revisited. J Fam Plann Reprod Health Care2009; 35: 211–3.8. Mackay J, et al., Schering Health Care Limited, OrganonLaboratories Limited, John Wyeth & Brother Limited.[2002] EWHC 1420(QB).9. Spitzer WO. The 1995 pill scare revisited: anatomy of anonepidemic. Hum Reprod 1997; 12: 2347–57.10. Ciliska D, Thomas H, Buffett C. An Introduction to Evi-dence-Informed Public Health and a Compendium of CriticalAppraisal Tools for Public Health Practice. February 2008.http://www.nccmt.ca/pubs/eiph_backgrounder.pdf [Accessed27 May 2010].11. American College of Obstetricians and Gynecologists(ACOG). Non-contraceptive Uses of Hormonal Contraception.ACOG Practice Guideline Number 110. January 2010. ObstetGynecol 2010; 115: 206–18.12. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O’FallonWM, Melton LJ 3rd. Trends in the incidence of deep veinthrombosis and pulmonary embolism. A 25 year populationbased study. Arch Int Med 1998; 158: 585–93.13. Levine JS, Branch DW, Rauch J. The antiphospholipid syn-drome. N Engl J Med 2002; 346: 752–63.14 Mohllajee AP, Curtis KM, Martins SL, Peterson HB. Doesuse of hormonal contraceptives among women with thrombo-genic mutations increase their risk of venous thromboembo-lism? A systematic review. Contraception 2006; 73: 166–78.

72 J Reproduktionsmed Endokrinol 2010; 7 (Special Issue 1)

Oral Contraceptives and Venous Thromboembolism

15. Pomp ER, Rosendaal FR, Doggen CJM. Smoking increasesthe risk of venous thrombosis and acts synergistically withoral contraceptive use. Am J Hematol 2008; 83: 97–102.

16. Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR,Colwell CW, et al. Prevention of venous thromboembolism:the Seventh ACCP Conference on Antithrombotic and Throm-bolytic Therapy. Chest 2004; 126: 338S–400S.

17. Chandra D, Parisini E, Mozaffarian D. Meta-analysis:travel and risk for venous thromboembolism. Ann Intern Med2009; 151: 180–90.

18. Pomp ER, le Cessie S, Rosendaal FR, Doggen CJR. Risk ofvenous thrombosis: obesity and its joint effect with oral con-traceptive use and prothrombotic mutations. Br J Haematol2007; 139: 289–96.

19. Ros HS, Lichtenstein P, Belloco R, Petersson G, CnattingiusS. Increased risks of circulatory diseases in late pregnancyand puerperium. Epidemiology 2001; 12: 456–60.

20. Pomp ER, Lenselink AM, Rosendaal FR, Doggen CJM. Preg-nancy, the postpartum period and prothrombotic defects: riskof venous thrombosis in the MEGA study. J Thromb Haemost2008; 6: 632–7.

21. Heinemann LAJ, Dinger JC. Range of published estimatesof venous thromboembolism incidence in young women. Con-traception 2007; 75: 328–36.

22. Dinger JC, Heinemann LAJ, Kuhl-Habich D. The safety of adrospirenone-containing oral contraceptive: final results fromthe European Active Surveillance study on Oral Contraceptivesbased on 142,475 women-years of observation. Contraception2007; 75: 344–54.

23. Heit JA, Kobbervig CE, James AH, Petterson TM, Bailey KR,Melton LJ 3rd. Trends in the incidence of venous thromboem-bolism during pregnancy or postpartum: a 30-year population-based study. Ann Intern Med 2005; 143: 697–706.

24. Ory HW. Mortality associated with fertility and fertilitycontrol: 1983. Fam Plann Perspect 1983; 15: 57–63.

25. Gerstman BB, Piper JM, Tomita DK, Ferguson WJ, StadelBV, Lundin FE. Oral contraceptive estrogen dose and the riskof deep venous thromboembolic disease. Am J Epidemiol 1991;133: 32–7.

26. Lidegaard Ø, Løkkegaard E, Svendsen AL, Agger C. Hormonalcontraception and risk of venous thromboembolism: nationalfollow-up study. BMJ 2009; 339: b2890.

27. Severinsen MT, Kristensen SR, Overvad K, Dethlefsen C,Tjønneland A, Johnsen SP. Venous thromboembolism discharge

diagnoses in the Danish National Patient Registry should beused with caution. J Clin Epidemiol 2010; 63: 223–8.

28. Gallo MF, Nanda K, Grimes DA, Schulz KF. Twenty micro-grams vs. > 20 microg estrogen oral contraceptives for contra-ception: systematic review of randomized controlled trials.Contraception 2005; 71: 162–9.

29. Culwell KR, Curtis KM. Use of contraceptive methods bywomen with current venous thrombosis on anticoagulanttherapy: a systematic review. Contraception 2009; 80: 337–45.

30. Canonico M, Plu-Bureau G, Lowe GD, Scarabin PY. Hormonereplacement therapy and risk of venous thromboembolism inpostmenopausal women: systematic review and meta-analysis.BMJ 2008; 336: 1227–31.

31. Jick SS, Hagberg KW, Hernandez RK, Kaye JA. Post-marketing study of ORTHO EVRA and levonorgestrel oral con-traceptives containing hormonal contraceptives with 30 mcgof ethinyl estradiol in relation to nonfatal venous thromboem-bolism. Contraception 2010; 81: 16–21.

32. De Lignieres B, Basdevant A, Thomas G, Thalabard JC,Mercier-Bodard C, Conard J, et al. Biological effects of estra-diol-17 beta in postmenopausal women: oral versus percuta-neous administration. J Clin Endocrinol Metab 1986; 62: 536–41.

33. Kroon UB, Silfverstolpe G, Tengborn L. The effects of trans-dermal estradiol and oral conjugated estrogens on haemosta-sis variables. Thromb Haemost 1994; 71: 420–3.

34. Alhenc-Gelas M, Plu-Bureau G, Guillonneau S, Kirzin JM,Aiach M, Ochat N, et al. Impact of progestagens on activatedprotein C (APC) resistance among users of oral contraceptives.J Thromb Haemost 2004; 2: 1594–600.

35. Fleming TR, DeMets DL. Surrogate end points in clinicaltrials: are we being misled? Ann Intern Med 1996; 125: 605–13.

36. Grimes DA, Schulz KF, Raymond EG. Surrogate end points inwomen’s health research: science, protoscience and pseudo-science. Fertil Steril 2010; 93: 1731–4.

37. Suissa S, Blais L, Spitzer WO, Cusson J, Lewis M, HeinemanL. First time use of newer oral contraceptives and the risk ofvenous thromboembolism. Contraception 1997; 56: 141–6.

38. Herings RM, Urquhart J, Leufkens HG. Venous thrombo-embolism among new users of different oral contraceptives.Lancet 1999; 354: 127–8.

39. Bloemenkamp KW, Rosendaal FR, Helmerhorst FM,Vandenbroucke JP. Higher risk of venous thrombosis during

early use of oral contraceptives in women with inherited clot-ting defects. Arch Intern Med 2000; 160: 49–52.

40. Cushman M, Kuller LH, Prentice R, Rodabough RJ, PsatyBM, Stafford RS, et al.; for the WHI Investigators. Estrogenplus progestin and the risk of venous thrombosis. JAMA 2004;292: 1573–80.

41. Seeger JD, Loughlin J, Eng PM, Clifford CR, Cutone J,Walker AM. Risk of thromboembolism in women taking ethi-nylestradiol/drospirenone and other oral contraceptives.Obstet Gynecol 2007; 110: 587–93.

42. Eng PM, Seeger JD, Loughlin J, Clifford CR, Mentor S,Walker AM. Supplementary data collection with case-cohortanalysis to address potential confounding in a cohort study ofthromboembolism in oral contraceptive initiators matched onclaims-based propensity scores. Pharmacoepidemiol Drug Saf2008; 17: 297–305.

43. Van Hylckama Vlieg A, Helmerhorst FM, VandenbrouckeJP, Doggen CJM, Rosendaal FR. Effects of oestrogen dose andprogestogen type on venous thrombotic risk associated withoral contraceptives: results of the MEGA case-control study.BMJ 2009; 339: b2921.

44. Shapiro S, Dinger J. Risk of venous thromboembolismamong users of oral contraceptives: a review of two recentlypublished studies. J Fam Plann Reprod Health Care 2010; 36:33–8.

45. US Food and Drug Administration (FDA). FDA monograph onphysician labelling for Yasmin 28 tablets (drospirenone and ethi-nyl estradiol). http://www.accessdata.fda.gov/drugsatfda_docs/label/2010/021098s017lbl.pdf [Accessed 27 May 2010].

46. Lidegaard O. Risk of venous thromboembolism among usersof oral contraceptives: a review of two recently published stud-ies (Letter). J Fam Plann Reprod Health Care 2010; 36: 103–4.

47. Shapiro S, Dinger J. Risk of venous thromboembolismamong users of oral contraceptives: a review of two recentlypublished studies (Letter). J Fam Plann Reprod Health Care2010; 36: 104–5.

48. Spitzer WO. Bias or causality: Interpreting recent evidenceof oral contraceptive studies. Am J Obstet Gynecol 1998; 179:S43–S50.

49. Shapiro S. Causation, bias and confounding: a hitchhiker’sguide to the epidemiological galaxy. Part 3: Principles of cau-sality in epidemiological research: statistical stability, dose-and duration-response effects, internal and external consis-tency, analogy and biological plausibility. J Fam Plann ReprodHealth Care 2008; 34: 261–4.

The Journal of Family Planning and Reproductive Health Care is the Journal of the Faculty of Sexual and ReproductiveHealthcare of the Royal College of Obstetricians and Gynaecologists. Visit the Faculty website at www.fsrh.org for furtherinformation about Membership or Associate Membership of the Faculty of Sexual and Reproductive Healthcare and fordetails of how to subscribe to the Journal (subscriptions to this quarterly journal include free online access).

Besuchen Sie unsere zeitschriftenübergreifende Datenbank

����� FallberichteP����� BilddatenbankP

����� ArtikeldatenbankP

e-journal-Abo

Beziehen Sie die elektronischen Ausgaben dieser Zeitschrift hier.Die Lieferung umfasst 6 Ausgaben pro Jahr zzgl. allfälliger Sonderhefte.Unsere e-Journale stehen als PDF-Datei (ca. 5–10 MB) zur Verfügung und sind auf denmeisten der marktüblichen e-Book-Readern, Tablets sowie auf iPad funktionsfähig.

����� Bestellung e-Journal-AboP

NEUES AUS DEM VERLAG

Besuchen Sie unsere Rubrik � � � � � Medizintechnik-Produkte

MediTEX IVFCritex GmbH

MEA-getesteteVerbrauchs-materialienGynemed GmbH

Zestica™Kairos LifeScience GmbH

InkubatorLabotect GmbH

Xario 200Toshiba MedicalSystems

Philips Clear Vue650 Mides GmbH

SteripetteMTG Medical

Seaforia™Origio GmbH

P

OvulaRingGynial GmbH