Color profile: Generic CMYK printer profile Composite Default screen

8

PATHOPHYSIOLOGY OF POSTPARTUM HEMORRHAGE AND THIRD STAGE OF LABOR R.-U. Khan and H. El-Refaey

INTRODUCTION uterine volume leads to a reduction in placental site surface area. As the is a relatively The physiology of postpartum hemostasis rigid and inelastic structure, the surface area of depends primarily upon mechanical events its attachment site decreases when it is tightly mediated by hormones, which induce strong compressed. uterine muscular contraction. Virtually all According to Brandt, compression of the recent studies focus on the latter, but the placenta forces placental back into the phenomenon cannot be understood without sinuses in the decidua basalis6. These sinuses examining why uterine contraction stops become blocked by the action of strong . Broadly speaking, and myometrial contraction, and thus the com- decidua are arranged such that powerful muscu- pressed placenta attempts to force blood back lar contractions after delivery favor hemostasis into a high-resistance system. Ultimately, the (Figure 1)1–3. Spiral arteries ‘fan out’ to create a sinuses become so congested that they rupture. low-resistance vascular bed in the intervillous The blood from the ruptured sinuses tears the space, which facilitates placental blood flow. fine septae of the spongy layer of the decidua This flow has been shown to decrease with basalis, and thus the placenta is sheared off7. muscular activity4. Third-stage contractions Dieckmann and colleagues implied that this are powerful and prolonged: they act to stop ‘retroplacental hematoma’ has no functional placental blood flow and to separate the value, and a subsequent investigation suggested placenta and membranes. that it is the contraction and retraction of the uterine wall itself that cause it to rend itself PLACENTAL SEPARATION AND apart from the placenta8. UTERINE ACTIVITY Ultrasonographic investigations recently corroborated that the Dieckmann theory is Mechanical events correct. Herman and colleagues conducted The biomechanical events which lead to real-time ultrasonographic imaging of the delivery of the placenta and its membranes third stage of labor and identified a ‘detach- begin to take place even before the start of the ment phase’, wherein the placenta completes second stage of labor. Membrane detachment its separation9. This detachment is preceded by starts during the first stage and slowly spreads a ‘contraction phase’, in which the placental- upwards from the internal os5. site uterine wall undergoes thickening. How- As the trunk of the baby is delivered, the ever, the ‘latent phase’ before this thickening muscle fibers undergo a very powerful occurred varied between patients and was contraction. Muscle fibers shorten, and the thought to determine the overall length of the uterus is reduced in size and volume, a process third stage. Of interest, neither the latent phase characterized as retraction. These events are nor the contraction phase was associated probably facilitated by the spiral arrangement of with ultrasound evidence of retroplacental uterine muscle fibers, whereby the reduction in hematoma formation.

62

84 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 30 August 2006 14:19:44 Color profile: Generic CMYK printer profile Composite Default screen

Pathophysiology of postpartum hemorrhage

membranes which follow the placenta and may form a retroplacental clot, whereas this blood escapes immediately in the Matthew Duncan method. Second, placental separation is slower in the Matthew Duncan method, allowing more time for bleeding10. As clinicians are able to nei- ther predict nor alter the method of placental separation, the distinction between the Schultze and Matthew Duncan methods is most proba- bly clinically irrelevant. Control of occurs by contraction and retraction of the interlacing myometrial fibers surrounding maternal spiral arteries of the placental bed. Myometrial con- traction compresses the spiral arteries and veins, thereby obliterating their lumina. It is for this reason that the myometrial fibers involved are often referred to as ‘living ligatures’10. In addi- tion, it is thought that some hemostasis occurs by means of direct pressure as the uterine walls are forced to firmly oppose one another as a result of myometrial contraction. It is worth noting the physiological effect of early cord clamping, a common intervention which is part of the active management of the third stage of labor, is to retain blood in the pla- centa, which prevents it from being so tightly Figure 1 (a) Circular uterine muscle at rest: two compressed by the uterus. This, in turn, reduces sets of crossing spiral; (b) at term: stretching of the the amount of myometrial retraction and con- 1 spirals (Goerttler, 1931 ). The innermost part of the traction, leading to more, not less, bleeding. muscular layer has been described as superficially However, this blood is thought to form a retro- ‘circular’ musculature, which is in fact two sets of placental clot, which speeds up the shearing off crossing spirals2. An alternative description of muscle fibers travelling in all directions has been of the placenta. Ultimately, the consequent described3. Both descriptions suggest that blood speedy delivery of placenta should lead to vessels are compressed during contraction of muscle quicker hemostasis, but the intervention of cord cells clamping is a paradox in that it involves causing increased initial bleeding to lessen ultimate total bleeding. The two classical methods of placental deliv- Unfortunately, apart from the recent ultra- ery result in different bleeding patterns. In the sound studies mentioned above, there is a dis- Schultze method, separation begins in the cen- tinct paucity of information about the physical ter of the placenta (the fetal surface), and this changes which lead to hemostasis and placental part descends first, with the remainder follow- separation. ing. The Matthew Duncan separation method involves detachment of the leading edge of the Endocrine mechanisms leading to placenta, and the entire organ slips down and mechanical events out of the uterus sideways. The latter method is much less common (20% of the total), but is Like all muscular activity, uterine contractility supposed to result in more bleeding for two pos- depends on both electrical and hormonal sible reasons. First, in the Schultze method, any stimuli. ‘Intrinsic’ activity may be mediated by extravasated blood is trapped within the stretch receptors, although it is unclear whether

63

85 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 30 August 2006 14:19:45 Color profile: Generic CMYK printer profile Composite Default screen

POSTPARTUM HEMORRHAGE

such mechanisms are neural or neurohormonal. both to prevent and to treat postpartum hemor- Two classes of hormones have been implicated rhage. At the same time, however, therapeutic in third-stage uterine contractility, namely oxytocic agents used to augment labor are and . sometimes associated with in the third stage. In this latter circumstance, the non-pulsatile administration of these agents Oxytocin may be leading to down-regulation of oxytocin Interest in the role of oxytocin in the third stage receptors, as has been demonstrated in in vitro has been partly motivated by the long-standing studies15. Despite the acknowledged therapeu- experience with therapeutic oxytocin to prevent tic role of oxytocic agents in the third stage of postpartum hemorrhage. Broadly speaking, labor, the true physiological role of oxytocin in oxytocin causes increased uterine contractions the third stage remains unclear. It appears to by acting on myometrial oxytocin receptors. have an inconsistent or paradoxical relationship However, research has failed to show a clear with the third stage. and simple relationship between physiological oxytocin action and third-stage events for a Prostaglandins number of reasons. Oxytocin assays are notori- ously unreliable, because the decidua synthe- Prostaglandins are potent stimulators of sizes its own oxytocin. As a result, plasma levels myometrial contractility, acting via cyclic AMP- do not reflect oxytocin concentrations at the mediated calcium release. The therapeutic use- myometrium. Moreover, plasma oxytocin levels fulness of agents in postpartum take no account of the density of myometrial hemorrhage lends credence to the possibility oxytocin receptors, which has been shown to of a physiological role for prostaglandins in participate in a complex control mechanism the third stage of labor. The prostaglandins with oxytocin itself and other factors. Finally, involved in uterine contraction are produced in oxytocinase, a plasma enzyme, denatures decidual tissue, placental tissue and fetal oxytocin before it reaches its site of action11. membranes16. The action of prosta- During labor, oxytocin is released in a glandins does not depend on gestation. There pulsatile manner, and both the pulse frequency are many classes of prostaglandin; the two and duration increase12. Exactly what triggers classes implicated in uterine contraction are the pulsatile oxytocin release is presently PGE2 and PGF2α. unclear. Ferguson speculated that uterine Several observers have noted that large stretching of the stimulates oxytocin amounts of prostaglandin are released in the release, leading to uterine contractions13. This third stage of labor. In an elegant experiment, phenomenon so far has not been demonstrated Noort and colleagues measured plasma levels of in humans, but there may be significant pres- prostaglandin metabolites during and up to 48 h 17 sure changes on adjacent pelvic organs and the after labor . PGF2α levels reached their maxi- vagina which result in neurological stimulation. mum and started to decline within 10 min after A pulse of oxytocin does not necessarily placental separation (Figure 2). The subsequent correspond to a uterine contraction, and some rapid decline in these levels suggested that the women do not experience a rise in plasma prostaglandins arise from either necrosis/cellu- oxytocin after the delivery of the baby14. More- lar disruption at the placental site, or from the over, it is not necessary to have an oxytocin fetal membranes. The latter are known to be a pulse in order to deliver the placenta and major source of prostaglandins. In vitro experi- achieve hemostasis. Additional methods of con- ments have shown that intrapartum amniotic trol must be involved. Whereas it is known that fluid triggers prostaglandin synthesis in fetal myometrial oxytocin receptor density increases membranes. The ‘active agent’ in the amniotic during and labor, the precise con- fluid remains unknown16; however, these obser- trols of this up-regulation are unknown15. vations are thought to reflect the active role of For many years, synthetic oxytocic agents prostaglandins in securing hemostasis by way of have been successfully used in the third stage myometrial contraction in the third stage.

64

86 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 30 August 2006 14:19:45 Color profile: Generic CMYK printer profile Composite Default screen

Pathophysiology of postpartum hemorrhage

1000

750 (pg/ml)

2 500 PGF

250

0 I II III 1 2 3 45 6 12 24 36 48 Time

Figure 2 Plasma PGF2α levels (pg/ml; mean ± standard equivalent of the mean). (I) In early labor and at full dilatation; (II) at delivery of the fetal head; (III) at placental separation and up to 48 h after placental separation (Noort et al., 198917)

The interaction between prostaglandins and Before and after delivery, subtle changes take endogenous or therapeutic oxytocin in the third place in both factors and fibrino- stage is not well understood. Numerous animal lysis agents. Plasma concentrations of clotting experiments have demonstrated interactions factors increase not only during pregnancy but between prostaglandins and oxytocin at luteo- also after delivery, which suggests a hyper- lysis, initiation and maintenance of pregnancy, coagulable state20. However, after placental and possibly at onset of labor18. However, ther- separation, the fibrinolytic potential of the apeutic oxytocic agents used in the third stage maternal blood also increases, and this tends to do not appear to have a significant effect on reduce the potential of blood to clot21. prostaglandin metabolite concentrations19. Fur- These conflicting changes are difficult to rec- ther studies are required to better understand oncile and are further complicated by changes from where the prostaglandins arise, and what in platelet activity before and after delivery. controls their release. However, there are indications that an inflam- In the next few years, it is likely that matory response arises at the placental bed after 22 , a analogue with placental delivery . Such a response would pro- uterotonic properties, will play an increasing mote local coagulation. This finding is impor- role in the management of the third stage, as it tant in terms of evolutionary advantage, because is both cheaper and more thermostable than it allows prevention of hemorrhage at the pla- existing agents. cental site, while elsewhere (particularly in deep pelvic and leg veins) thrombi are less likely to persist, due to the increased fibrinolysis. Coagulation von Willebrand disease (factor VIII defi- Many standard obstetric text books provide ciency) is an important example of a coagulo- only the vaguest of suggestions that coagulation pathy which can result in increased risk of at the placental site represents an important postpartum hemorrhage. This is especially true hemostatic mechanism. Whilst this is certainly in the disease variant featuring factor VIIIc defi- true, the exact pathway(s) involved are unclear. ciency. In many ways, von Willebrand disease

65

87 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 04 September 2006 13:40:59 Color profile: Generic CMYK printer profile Composite Default screen

POSTPARTUM HEMORRHAGE

mimics a platelet adhesion dysfunction, and time. Whilst uterine atony is responsible for indeed the only aspect of hematological 75–90% of primary postpartum hemorrhage, hemostasis after placental delivery which can be traumatic causes of primary postpartum hemor- emphasized with any certainty is the formation rhage (including obstetric lacerations, uterine of platelet plugs at arterioles. Postpartum inversion and ) comprise about hemorrhage rates in von Willebrand’s disease 20% of all primary postpartum hemorrhage (see are in excess of 15%, and it has been suggested Chapter 9). Significant but less common causes that this hemorrhage is largely preventable by of postpartum hemorrhage include congenital minimizing maternal trauma at delivery and giv- and acquired clotting abnormalities, which ing prophylactic treatment with desmopressin comprise around 3% of the total24. Uterine (DDAVP)23. atony is responsible for the majority of primary In summary, the hemostatic mechanisms postpartum hemorrhage originating from the during and after placental separation probably placental bed. Although the most important involve the contraction of muscle sheaths around risk factor is a previous history of atonic post- the spiral arteries, leading to platelet plug forma- partum hemorrhage (relative risk 3.3)25, many tion, retraction of the uterus causing mechanical other important risk factors often found in occlusion of arterioles facilitating platelet plug combination. formation, and the activation of both the clotting Failure of the uterus to contract may be cascade and fibrinolysis. As of this writing, many associated with or placental of these events are vague assumptions rather than fragments, either as disrupted portions, or more demonstrated fact, as third-stage physiology rarely a succenturiate lobe. The retained mate- research has been grossly neglected. The fact rial acts as a physical block against strong uter- that for decades effective treatments have been ine contraction, which is needed to constrict available for postpartum hemorrhage in the placental bed vessels, but, in most cases, dys- developed world has acted as a true disincentive functional postpartum contraction is the pri- for novel work and ideas. It is tragic that the third mary reason for placental retention. It is more stage of labor, the most dangerous moment of likely for the placenta to be retained in cases of pregnancy, is so poorly understood. atonic postpartum hemorrhage, and so the con- traction failure often becomes self-perpetuating. The reasons for this contractile dysfunction are PATHOPHYSIOLOGY OF unknown. The exception is uterine fibroids, POSTPARTUM HEMORRHAGE where the source of distension cannot be Although most of the physiological processes in removed by uterine contraction, and must the third stage of labor remain unclear, they therefore cause the atony. However, the uterus broadly help to explain the etiology of atonic does not even have to be distended during the postpartum hemorrhage. In this section, the third stage for contractile dysfunction to occur. etiology and accompanying pathophysiology Distension prior to delivery, which occurs with will be discussed. multiple pregnancy and , also affects the ability of the uterus to contract efficiently after delivery, and is thus another Uterine atony risk factor for atonic postpartum hemorrhage. The most common cause of postpartum hemor- When postpartum hemorrhage occurs fol- rhage is uterine atony, i.e. failure of the uterus lowing an antepartum hemorrhage, the scenario to contract. Primary postpartum hemorrhage is particularly difficult since there have been two due to uterine atony occurs when the relaxed episodes of blood loss. A rare but serious com- myometrium fails to constrict these blood plication of abruption is extravasation of blood vessels, thereby allowing hemorrhage. Since into the myometrium, known as a Couvelaire up to one-fifth of maternal cardiac output, or uterus, which impairs the physiological uterine 1000 ml/min, enters the uteroplacental circula- contraction/retraction hemostatic process. tion at term, postpartum hemorrhage is capable However, the relationship between the extra- of exsanguinating the mother within a short vasation process and uterine dysfunction is

66

88 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 30 August 2006 14:19:50 Color profile: Generic CMYK printer profile Composite Default screen

Pathophysiology of postpartum hemorrhage

not fully understood. has a the task of postpartum contraction/retraction, similar effect for unknown reasons. Both ante- and thus hemorrhage ensues26. As contraction/ partum hemorrhage and chorioamnionitis also retraction are considered essential prerequisites impair uterine contraction during the first two for both placental detachment and postpartum stages of labor, and prolonged labor in general is hemostasis, the inference is that physiological a risk factor for postpartum hemorrhage. Con- hemostasis from a lower segment placental bed ventional wisdom suggests that delay in the first is impossible. This is obviously not the case, two stages leads to uterine atony, but it is more however, as clearly not all cases of Grade IV pla- logical to suggest that uterine dysfunction centa previa necessitate . The only before onset of labor results in delay in all three possible conclusion is that there are qualitative stages, and thus causes postpartum hemor- and quantitative differences in the musculature rhage. As far as we are aware, there is no of the lower segment in different patients. A ongoing research into this ‘universal uterine recent literature search on this topic confirms dysfunction’. that the nature and origin of these differences have never been investigated. Biswas and colleagues have compared pla- The lower segment as an implantation site cental bed biopsy changes in placenta previa In both placenta previa and placenta accreta, and normally implanted placenta; they have the placental bed (and thus the postpartum shown that previa is associated with significantly bleeding site) is in the lower segment. The pres- higher trophoblastic giant cell infiltration and ence of lower segment implantation makes physiological changes of the myometrial spiral hemorrhage and placental retention much more arterioles27. This work is typical of modern likely. Although existing evidence is scanty, obstetric research in that it concentrates on there are indications that the etiology of patho- antenatal events while ignoring postpartum logical bleeding is inextricably linked with the events. However, the findings are interesting anatomical and physiological limitations of the because they suggest that the seeds of potential lower segment. placenta accreta are sown in most cases of placenta previa. Nonetheless, no knowledge regarding the qualitative features of lower Placenta previa segment myometrium exists. In placenta previa, the placental site is located in an abnormally low position. Atonic postpartum Placenta accreta hemorrhage is a recognized and, even if Cesarean section is performed, severe Placenta accreta is morbid adherence of intraoperative bleeding is a significant risk26. placenta such that it invades the myometrium. The usual pharmacological methods used to It is rare; in 1990, the quoted incidence was stem hemorrhage are often less effective. Surgi- around 1 in 2000 to 1 in 3500 pregnant women cal methods, such as oversewing of bleeding in North America28. The condition is strongly sinuses and the B-Lynch suture, are sometimes linked to lower segment implantation; it occurs also ineffective so that hysterectomy proves nec- in up to 15% of women with placenta previa26. essary. Hemorrhage is often not stopped unless The adherence is also associated with a the entire lower segment is removed; a subtotal deficiency of decidua in the lower segment. hysterectomy is often inadequate, and many The most common cause of this decidual surgeons perform total abdominal hysterectomy deficiency is endometrial scarring, which may as the operation of choice. Thus, the involve- be secondary to previous Cesarean section ment of the lower segment makes it more likely or myomectomy, previous endometritis, past not only that hemorrhage will occur, but also history of evacuation of retained products of that standard treatment modalities will fail. conception or uterine abnormalities. Authors in conventional texts often suggest Uterine surgery is a major risk factor for pla- that, in lower segment implantation, the muscle centa previa and placenta accreta29. There is an surrounding the placental bed is inadequate to increased tendency for placental implantation in

67

89 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 30 August 2006 14:19:50 Color profile: Generic CMYK printer profile Composite Default screen

POSTPARTUM HEMORRHAGE

the vicinity of the uterine scar with secondary because they tend to be relatively less trophoblast invasion of the myometrium. Uter- amenable to medical treatment and sometimes ine scarring is also known to be associated with necessitate radical surgical intervention, such as an increased risk of scar dehiscence, febrile mor- hysterectomy. bidity and other factors30. Thus the scar is clas- Whereas knowledge of the ultrastructure sically considered to be a ‘weak area’. Scarring of placental bed musculature is lacking with of muscle results in the normal tissue being regards to the upper segment, it is virtually non- replaced by fibrous tissue. Intrauterine retrac- existent for the lower segment. New research tion forces induced during labor tend to thin out into this area is urgently needed, because all the lower segment, and these forces stretch the non-surgical therapeutic modalities for post- scar to the point of rupture. Uterine rupture is partum hemorrhage involve enhancement of not considered predictable31, but is more likely uterotonicity and, in the absence of sufficient with each Cesarean section. Although poorly myometrium, they will simply not work. We described in the literature, our personal clinical hypothesize that lower segment placentation/ experience suggests that, with each ensuing surgery leads to structural and thus functional Cesarean section, the entire lower segment changes in the muscle histology. Thus, we often seems to become thinner. Indeed, the envisage a new, clinically important class lower segment may take on a translucent of postpartum hemorrhage, ‘lower segment quality. This appearance is not limited to the postpartum hemorrhage’. This new subclass scar itself. It is possible that the ‘weak scar’ will be best managed by new protocols which in fact represents a generalized lower segment address the features specific to lower segment weakness induced by previous surgery. involvement. Clinical experience also suggests to us that it is not enough to assume that postpartum hem- orrhage is more common with lower segment References implantation purely because lower segment 1. Goerttler K. Die Architektur der Muskelwand muscle is inadequate to the task. In cases of des menschlichen Uterus ind ihre funktionelle placenta previa and placenta accreta, the lower Bedeutung. [The architecture of the muscle segment looks even thinner than normal. We bonds of the human uterus and their functional hypothesize that the contractile nature of lower behavior.] Gegenbaurs morphologisches Jahrbuch segment muscle, which is already less than that 1931;45–128 of the upper segment, is further lowered by the 2. Fuchs A, Fuchs F. Physiology of parturition. In presence of the placenta. This would mean Gabbe S, Niebyl J, Simpson J, eds. : that implantation itself has an adverse effect Normal and Problem , 2nd edn. New York: Churchill Livingstone, 1991:147–74 on lower segment myometrium. Furthermore, 3. Renn K. Untersuchungen ueber die raeumliche there is a body of anecdotal evidence which Anordnung der Muskelbuendel im Corpus implies that placental size and trophoblast bereich des menschilichen Uterus. Z Anat invasion are greater in areas of limited decidual Entwicklungsgesch 1970;132:75–106 tissue, including implantation on scars and 4. Lees M, Hill J, Ochsner A, et al. Maternal pla- in ectopic pregnancies. We hypothesize that cental and myometrial blood flow of the rhesus trophoblast would invade readily into the poorly monkey during uterine contractions. Am J Obstet decidualized lower uterine segment, increasing Gynecol 1971;110:68–81 the likelihood that placenta accreta will develop. 5. de Groot A. Safe motherhood – the role of oral In terms of the previous discussion, it is (methyl) in the prevention of unfortunate that a dramatic and remorseless rise postpartum hemorrhage. MD Thesis, University of Nijmegen, 1995 in the Cesarean section rate is being observed 6. Brandt M. The mechanism and management of throughout the developed world. This phenom- the third stage of labor. Am J Obstet Gynecol enon will inevitably give rise to an increase 1933;25:662–7 in the complications associated with placenta 7. Dieckmann W, Odell L, Williger V, et al. The previa, placenta accreta and scar rupture. placental stage and postpartum hemorrhage. Am The complications are particularly important J Obstet Gynecol 1947;54:415–27

68

90 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 30 August 2006 14:19:51 Color profile: Generic CMYK printer profile Composite Default screen

Pathophysiology of postpartum hemorrhage

8. Inch S. Management of the third stage of labour pregnancy and puerperium. Thromb Res 1989; – another cascade of intervention? Midwifery 54:91–8 1985;1:114–22 22. Louden K, Broughton Pipkin F, Symonds F, 9. Herman A, Weinrauth Z, Bukovsky I, et al. et al. A randomised placebo-controlled study of Dynamic ultrasonographic imaging of the third the effect of low dose aspirin on platelet reactivity stage of labor. New perspectives into third stage and serum thromboxane B2 production in non- mechanisms. Am J Obstet Gynecol 1993;168: pregnant women, in normal pregnancy, and in 1496–9 . Br J Obstet Gynaecol 10. Sweet D, Kiran B. Mayes’ Midwifery. London: 1992;99:371–6 Balliere Tindall, 1997 23. Kadir R, Lee C, Sabin C, et al. Pregnancy in 11. Hirst J, Chibbar R, Mitchell B. Role of oxytocin women with von Willebrand’s disease or factor in the regulation of uterine activity during XI deficiency. Br J Obstet Gynaecol 1998;105: pregnancy and in the initiation of labour. Semin 314–21 Reprod Endocrinol 1993;11:219–33 24. Prendiville W, Elbourne D. Care during the 12. Fuchs A, Romero R, Keefe D, et al. Oxy- third stage of labour. In Chambers I, Enkin M, tocin secretion and human parturition: pulse Keirse M, eds. Effective Care in Pregnancy and frequency and duration increase during , Vol 2. Oxford: Oxford University spontaneous labour in women. Obstet Gynecol Press, 1989:1145–70 1991;165:1515–23 25. Stones R, Paterson C, Saunders N. Risk factors 13. Ferguson J. A study of the motility of the intact for major obstetric hemorrhage. Eur J Obstet uterus of the rabbit at term. Surg Gynecol Obstet Gynecol Reprod Biol 1993;48:15–18 1941;73:359–66 26. Konje J, Whalley R. Bleeding in late pregnancy. 14. Thornton S, Davison J, Baylis P. Plasma In James D, Steer P, Weiner C, Gonik B, oxytocin during third stage of labour: compari- eds. High-risk Pregnancy Management Options. son of natural and active management. Br Med J London: Saunders, 1994:119–36 1988;297:167–9 27. Biswas R, Sawhney H, Dass R, Saran R, Vasishta 15. Phaneuf S, Asboth G, Carrasco M, et al. Desen- K. Histopathological study of placental bed sitisation of oxytocin receptors in human myo- biopsy in placenta previa. Acta Obstet Gynecol metrium. Hum Reprod Update 1998;4:625–33 Scand 1999;78:173–9 16. Brennand J, Leask R, Kelly R, et al. The influ- 28. Zahan C, Yeomans E. Postpartum hemorrhage: ence of on prostaglandin synthesis placenta accreta, and puerperal and metabolism in human fetal membranes. Acta hematomas. Clin Obstet Gynecol 1990;33:422–31 Obstet Gynecol Scand 1998;77:142–50 29. Dickinson J. Previous . In 17. Noort W, van Buick B, Vereecken A, et al. James D, Steer P, Weiner C, Gonik B, eds. Changes in prostaglandin levels of PGF2α and High-risk Pregnancy Management Options. PGI2 metabolites at and after delivery at term. London: Saunders, 1994;207–16 Prostaglandins 1989;37:3–12 30. Enkin M, Wilkinson C. Manual removal of 18. Jenkin G. Oxytocin and prostaglandin inter- placenta at caesarean section (Cochrane review). actions in pregnancy and at parturition. J Reprod In Keirse MJNC, Renfrew MJ, Neilson JP, Fertil 1992;45(Suppl):97–111 Crowther C, eds. Pregnancy and Childbirth 19. Ilancheran A, Ratnam S. Effect of oxytocics on Module. In The Cochrane Pregnancy and Child- prostaglandin levels in the third stage of labour. birth Database [database on disk and CDROM]. Gynecol Obstet Invest 1990;29:177–80 The Cochrane Collaboration; Issue 2, Oxford: 20. Wallenburg H. Changes in the coagulation Update Software, 1999. Available from BMJ system and platelets in pregnancy-induced Publishing Group, London hypertension and pre-. In Sharp F, 31. Beasley J. Complications of the third stage of Symonds E, eds. Hypertension in Pregnancy. labour. In Whitfield C, ed. Dewhurst’s Textbook Ithaca: Perinatology Press, 1987:227–48 of Obstetrics and Gynaecology for Postgraduates, 21. Shimada H, Takshima E, Soma M, et al. Source 5th edn. Oxford: Blackwall Science 1995: of increased plasminogen activators during 368–76

69

91 Z:\Sapiens Publishing\A5211 - Postpartum Hemorrhage\Make-up\Postpartum Hemorrhage - Voucher Proofs #T.vp 30 August 2006 14:19:51