- Email: [email protected]
Antenatal fibrinogen concentrations and postpartum haemorrhage
International Journal of Obstetric Anesthesia (2014) 23, 365–370 0959-289X/$ - see front matter c 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijoa.2014.06.004
ORIGINAL ARTICLE
www.obstetanesthesia.com
Antenatal fibrinogen concentrations and postpartum haemorrhage T. Yamada, R. Akaishi, Y. Oda, R. Nishida, T. Yamada, S. Ishikawa, M. Morikawa, T. Kojima, H. Minakami Department of Obstetrics, Hokkaido University Graduate School of Medicine, Sapporo, Hokkaido, Japan ABSTRACT Background: It is unclear whether antenatal fibrinogen concentrations are associated with postpartum haemorrhage. Methods: This retrospective study included 871 women with a singleton pregnancy but no known risk factors for postpartum haemorrhage, in whom fibrinogen concentration was measured within the 21 days before delivery. Correlation between antenatal fibrinogen concentrations and estimated blood loss was analysed. We tested the hypothesis that the risk of postpartum haemorrhage was higher in women with antenatal fibrinogen concentrations of <3.3 g/L. Postpartum haemorrhage was defined as an estimated blood loss P700 mL following vaginal delivery and P1000 mL following caesarean delivery. Results: In women delivering vaginally (n = 337), estimated blood loss tended to increase with decreasing antenatal fibrinogen concentration (R = 0.107, P = 0.05), median fibrinogen concentration was significantly lower in 69 women with postpartum haemorrhage than in 268 women without postpartum haemorrhage (3.93 vs. 4.18 g/L, P = 0.025), and postpartum haemorrhage occurred significantly more often in women with fibrinogen concentrations <3.3 g/L than in those with concentrations P3.3 g/L (38% [11/29] vs. 19% [58/308], P = 0.018). In women undergoing caesarean delivery (n = 534), median fibrinogen concentration did not differ between those who experienced postpartum haemorrhage (n = 128) and those who did not (n = 406) (4.18 g/L vs. 4.07 g/L, P = 0.43). Antenatal fibrinogen concentrations of <3.3 g/L were not associated with higher rates of postpartum haemorrhage (26% [11/43] vs. 24% [117/491], P = 0.80). Conclusions: Antenatal fibrinogen concentration <3.3 g/L may be a risk factor for postpartum haemorrhage among women following vaginal delivery. c 2014 Elsevier Ltd. All rights reserved.
Keywords: Postpartum haemorrhage; Fibrinogen; Vaginal delivery; Caesarean delivery
Introduction Postpartum haemorrhage (PPH) is the leading cause of maternal mortality, accounting for approximately one third of maternal deaths in Africa and Asia,1 and it remains a major cause of maternal mortality in developed countries. PPH has been associated with 12.5–19.1% of maternal mortality in the USA.2,3 In Japan, the maternal mortality rate is <5 per 100 000 births according to the Japanese Ministry of Heath, Labour and Welfare4 with PPH accounting for 14.1%, 9.7% and 6.7% of all causes of maternal mortality in Japan in 2000, 2005 and 2010, respectively.5 Although risk factors such as prolonged, augmented or rapid labour, history of PPH, preeclampsia, distended uterus and operative delivery are associated Accepted June 2014 Correspondence to: Takashi Yamada, Department of Obstetrics, Hokkaido University Graduate School of Medicine, Kita-ku N14 W6, Sapporo 060-8638, Japan. E-mail address: [email protected]
with PPH,6 it can occur without warning. Causes such as uterine rupture, uterine inversion, retained placenta and birth canal injury may be identified; however, the leading aetiology of PPH is uterine atony, and many women who develop PPH from uterine atony do not have known risk factors.2 There is interest in how the maternal coagulation profile alters during the course of PPH and fibrinogen has been identified an important factor that may influence the magnitude of blood loss.7–11 Studies have demonstrated that women who develop severe PPH have a low fibrinogen concentration in the initial stage of haemorrhage, that fibrinogen concentration in the initial stages of haemorrhage is a significant predictor of severity of PPH, and that decreases in fibrinogen concentration are closely correlated with the severity of blood loss. These studies have not determined whether a low fibrinogen concentration in the initial stages of PPH is caused by haemorrhage itself or reflects fibrinogen concentration before labour. It is unclear whether a low antenatal fibrinogen concentration is associated with PPH.
366
Antenatal fibrinogen and postpartum haemorrhage
This retrospective study was conducted to determine whether women with lower antenatal fibrinogen concentrations had a higher risk of developing PPH in the absence of known risk factors for PPH.
Methods This retrospective study was conducted with the approval of the Ethics Committee of Hokkaido University Hospital. The institutional delivery database was searched for women who gave birth to a singleton infant at or after gestational week 32 at Hokkaido University Hospital between 1 April 2007 and 31 March 2013. Women who had their fibrinogen levels checked in the 21 days pre-delivery and who had no known risk factors for PPH were included. Women were divided into two groups according to whether they had a vaginal or caesarean delivery. PPH was defined as an estimated blood loss (EBL) P700 mL for vaginal delivery and P1000 mL for caesarean delivery. Laboratory tests, including full blood count, biochemistry and coagulation-fibrinolysis are performed for the majority of pregnant women at our clinic and for those admitted to hospital for management of obstetric and incidental complications. All blood fibrinogen data from the institutional central laboratory were obtained for each patient during the study period. In some women, fibrinogen concentration was measured several times within 21 days of delivery. The fibrinogen value determined on the day closest to delivery was used. Demographic characteristics and EBL within 24 h of delivery were recorded from medical charts. Fibrinogen concentration was measured using the Clauss thrombin clotting time method in citrated blood samples and a haematology analyzer (CS2000i, Sysmex Co. Ltd., Kobe, Japan) after centrifugation. Intra- and inter-assay coefficients of variation were <6% and 10%, respectively. Normal fibrinogen concentration at our institution, not specific for pregnancy, is 2–4 g/L.
Statistical analysis Linear regression analysis was used to test the hypothesis that the risk of PPH was higher in women with antenatal fibrinogen concentration of <3.3 g/L than in those with antenatal fibrinogen concentration of P3.3 g/L. Data are presented as median [range] or mean ± standard deviation (SD). Statistical analyses were performed using the JMP10 c statistical software package (SAS, Cary, NC, USA). ANOVA and Tukey–Kramer HSD (honestly significant difference) tests were used for comparison of means. The Wilcoxon/Kruskal–Wallis method was used for comparison of medians. Fisher’s exact test was used for comparison of categorical variables. In all analyses, P < 0.05 was considered statistically significant.
Results A total of 1473 women were identified: 766 following vaginal delivery and 707 after caesarean delivery (Fig. 1). After matching fibrinogen data with medical charts, 925 women were found to have had blood fibrinogen concentrations measured in the 21 days before delivery. Fifty-four women with known risk factors for PPH before or after delivery were excluded: 52 with placenta praevia or low-lying placenta with or without placenta accreta, one who underwent radical hysterectomy after caesarean delivery for treatment of uterine cervical cancer, and one with uterine inversion immediately after delivery. The study therefore looked at data from 871 women, representing 59% of the 1473 women originally identified. Of the 871 women, 337 delivered vaginally and 534 underwent caesarean delivery (Table 1). For the 337 women delivering vaginally, EBL within 24 h after delivery was P700 mL in 67 (20%), >600 mL in 94 (28%) and >500 mL in 142 (42%). For caesarean deliveries, EBL within 24 h after delivery was P1000 mL in 128 of 534 women (24%). In women delivering vaginally, median [range] fibrinogen concentration of those who had a PPH was significantly lower than that of women who did not (3.93 [2.57–5.71] g/L vs. 4.18 [2.08–6.30] g/L, P = 0.025). For caesarean delivery, there was no difference in median [range] fibrinogen concentration for 128 women who had a PPH compared to the 406 who did not (4.18 [2.57–6.66] g/L vs. 4.07 [1.79–8.98] g/L, P = 0.43). The distribution of fibrinogen concentrations is shown in Fig. 2. Distribution did not differ between women in the vaginal and caesarean delivery groups: mean fibrinogen concentrations for women delivering vaginally or by caesarean were similar (4.13 ± 6.7 g/L vs. 4.17 ± 7.0 g/L, P > 0.05). Fibrinogen concentrations of <3.0 g/L and <3.3 g/L were detected in 23 (2.6%) and 72 (8.3%) of all patients, respectively. Estimated blood loss tended to increase with decreasing antenatal fibrinogen concentration in 337 women with vaginal deliveries (R = 0.1067, P = 0.05) (Fig. 3), but not in 534 women with caesarean deliveries (R = 0.048, P = 0.26). Among women delivering vaginally, the frequency of PPH was significantly higher in women with fibrinogen concentrations of <3.3 g/L than in those with fibrinogen concentrations of P3.3 g/L (Fig. 4). A fibrinogen cut-off value of 3.3 g/L yielded sensitivity of 16% (11/ 69), specificity of 93% (250/268), positive predictive value of 38% (11/29), and negative predictive value of 81% (250/308) for PPH. In women undergoing caesarean delivery, the frequency of PPH did not differ significantly for antenatal fibrinogen concentration above or below 3.3 g/L.
T. Yamada et al.
367 Number of women with singleton pregnancies who gave birth at ≥ 32 weeks of gestation during the study period
1473 women (766 vaginal and 707 caesarean deliveries)
Available fibrinogen level within 21 days pre-delivery?
No (n=548)
925 women (338 vaginal and 587 caesarean deliveries)
Risk factors for postpartum haemorrhage?
Yes (n=54)
871 women (337 vaginal and 534 caesarean deliveries) Fig. 1
Table 1
Flow diagram showing criteria for selection
Patient characteristics
Age (years) Nulliparous Body mass index (kg/m2) Gestation (weeks) Gestation <37 weeks Fibrinogen (g/L) Estimated blood loss P500 mL P600 mL P700 mL P1000 mL Birthweight (<2500 g) Birthweight <2500 g
Vaginal delivery (n = 337)
Caesarean delivery (n = 534)
P value
32 [18–44] 198 (58.8%) 20.1 [14.1–41.9] 38 [32–42] 52 (15.4%) 4.11 [2.08–6.3]
33 [18–48] 278 (52.1%) 21.2 [15.0–46.6] 37 [32–42] 88 (16.5%) 4.09 [1.79–8.98]
<0.0001 0.053 <0.0001 <0.0001 0.68 0.71
142 (42.1%) 96 (28.5%) 69 (20.5%) 28 (8.3%) 2925 [1260–4420] 69 (20.5%)
413 (78.3%) 345 (64.6%) 271 (50.7%) 128 (24%) 2815 [1102–5586] 127 (23.8%)
<0.0001 <0.0001 <0.0001 <0.0001 0.0028 0.25
Data are median [range] or number (%).
Discussion This study demonstrated that women with lower antenatal fibrinogen concentrations who gave birth vaginally were at higher risk of PPH; this was not seen in women undergoing caesarean delivery. To our knowledge, this has been noted in only one previous study.12 Japanese Guidelines for Obstetrical Practice13 report a 90th percentile value of EBL after vaginal delivery of 800 mL. In this study, EBL P500 mL, P600 mL and P700 mL occurred in 42%, 28% and 20% respectively, of the 337 women with vaginal delivery, findings consis-
tent with local recommendations. In addition, since only 0.3% of Japanese women require blood transfusion for excessive blood loss after vaginal delivery,14 it may be reasonable to define PPH following vaginal delivery as an EBL of P700 mL for Japanese women. Fibrinogen concentrations <3 g/L and <3.3 g/L were seen in 2.6% and 8.3% of all patients in the study. Simon et al. reported that 4.1% of 677 women delivering vaginally had an antenatal fibrinogen concentration <2.9 g/L. PPH, defined as clinically severe bleeding requiring volume expansion and manual uterine exploration, associated with either blood transfusion or a decrease in haemoglobin concentration P2.0 g/
Antenatal fibrinogen and postpartum haemorrhage Percentage (number) of women
368 35
(265)
30
(243)
25 20
(138) 15
(110)
10 5
(60) (32)
(23)
0 <3 -299
3-3.49 350-399 3.5-3.99 400-449 4-4.49 450-499 4.5-4.99500-549 5-5.49 550>5.5 300-349
Fibrinogen level (g/L)
Fig. 2 Fibrinogen concentrations in the 21 days before delivery
n=337
Estimated blood loss (mL)
R=−0.11 P=0.0504
2
3 4 5 6 Antenatal fibrinogen level (g/L)
Fig. 3 Correlation between antenatal fibrinogen concentrations and estimated blood loss
P=0.018
Percentage of women with postpartum haemorrhage
40
(11/29)
30 20
(11/43)
(117/491)
(58/308)
10 0
<3.3g/L ≥ 3.3g/L Vaginal delivery
<3.3g/L ≥ 3.3g/L Caesarean delivery
Fig. 4 Frequency of postpartum haemorrhage in vaginal and caesarean delivery groups analysed using an antenatal fibrinogen cut-off of 3.3 g/L
dL, or both, within 24 h of delivery occurred in 6.8% of the 677 women. A cut-off fibrinogen value of 2.9 g/L yielded a sensitivity of 19.6% and a positive-predictive value of 39.1% for PPH.12 Using our criteria for PPH of an EBL P700 mL, the cut-off of 3.3 g/L yielded sensitivity of 16% and positive-predictive value of 38% for prediction of PPH following vaginal delivery. Peyvandi et al. examined the risk of PPH according to increasing antenatal fibrinogen concentration in 4461 women and failed to show any difference in five groups (fibrinogen concentration 64 g/L, 4.01–4.5 g/L, 4.51–5 g/L, 5.01–5.50 g/L and >5.51 g/L).15 In the present study, the risk of PPH did not differ significantly between women with fibrinogen levels above or below 3.6 g/L (data not shown), consistent with the results reported by Peyvandi et al. However, an antenatal fibrinogen concentration of 4 g/L was the 20th percentile value in Peyvandi’s study, while in our study it corresponded to the 43rd percentile value, and our 20th percentile fibrinogen value was 3.6 g/L. The reason for this difference is not clear. Fibrinogen concentration increases during pregnancy from a non-pregnant concentration of 2.52 ± 0.7 g/L to 4.47 ± 0.9 g/L by 33 weeks of gestation. 16 Fibrinogen is consumed at the time of caesarean delivery. Uterine vein fibrinogen concentration immediately before uterine wall incision falls from 4.25 ± 0.21 to 4.1 ± 0.2 g/L and 4.05 ± 0.22 g/L, 5 and 15 min after removal of the placenta. Fibrinogen concentration in the uterine vein after removal of the placenta is significantly lower than that in the circulating blood,17 and its concentration in circulating blood decreases from approximately 4.0 g/L to 3.46 g/L within 4 h of caesarean delivery.18 Current guidelines recommend correcting fibrinogen concentration if <1.0 g/L,19,20 although the rate of clot formation is not optimised until fibrinogen concentrations reach >2.0 g/L in an in vitro haemodilution model.21 Studies have suggested that the fibrinogen concentration at an early stage of PPH is associated with severity of haemorrhage,7,10 and a lower fibrinogen concentration is likely to occur with increasing blood loss.9 The risk of severe PPH has been shown to be 2.63-fold higher for each 1 g/L decrease in fibrinogen concentration at the beginning of PPH.7 Mean fibrinogen concentration at diagnosis of PPH has been shown to be significantly lower in women with subsequent worsening of PPH compared to that in women without worsening of PPH (3.4 ± 0.9 g/L vs. 4.2 ± 1.2 g/L).10 The lowest fibrinogen concentration measured within 24 h of delivery has the best correlation with increasing severity of PPH.8 In one study, median fibrinogen concentration on admission for treatment of PPH was significantly lower in women requiring intervention, such as uterine artery embolisation and/or haemostatic surgery, than in those who did not undergo such procedures (1.8 g/L vs. 2.5 g/L, P < 0.0001).9
T. Yamada et al. Women undergoing caesarean delivery who had low antenatal fibrinogen concentration did not have a higher risk of PPH in this study. Uterine atony is a leading cause of PPH after vaginal delivery.2,22 Although uterine atony is ill-defined, insufficient uterine involution is thought to lead to PPH. Uterine atony is responsible for 79% of PPH events in the USA.2 Continuous uterine massage facilitates involution and is recommended for the treatment of PPH, but not as an intervention to prevent PPH after vaginal delivery in women who have received prophylactic oxytocin because uterine massage may cause maternal discomfort.22 Our routine procedures include continuous uterine massage during caesarean delivery, but not after vaginal delivery. This may have reduced the effect on PPH of low antenatal fibrinogen concentration following caesarean delivery in the present study. The study has some limitations. Selection bias is a potential concern. There were 766 vaginal deliveries during the study period, but only 337 women (44%) were included in the study; this was mainly due to limited or absent fibrinogen data within the 21 days before delivery. This suggests that the final three weeks of pregnancy were uneventful in 56% of women excluded from the present analysis. However, our protocol for management of pregnant women did not include treatment options that were determined by antenatal fibrinogen concentration. It seems reasonable to speculate that obstetric management did not differ from those women in whom the fibrinogen concentration was unknown. Many clinical factors may affect blood loss after vaginal delivery, and it was not clear from this study whether a low fibrinogen concentration was an independent risk factor for PPH. In conclusion, the present study suggests that women who deliver vaginally with a relatively low antenatal fibrinogen concentration had a higher risk of developing PPH. However, it remains unclear whether antenatal low fibrinogen concentration was an independent risk factor for PPH. Larger prospective studies are needed to clarify this issue.
Disclosure This study was supported by departmental funds only and the authors have no conflicts of interest to declare.
Acknowledgement We thank the staff of the Department of Obstetrics, Hokkaido University Hospital, for their cooperation in blood sampling.
369
References 1. Khan KS, Wojdyla D, Say L, Gu¨lmezoglu AM, Van Look PF. WHO analysis of causes of maternal death: a systematic review. Lancet 2006;367:1066–74. 2. Bateman BT, Berman MF, Riley LE, Leffert LR. The epidemiology of postpartum hemorrhage in a large, nationwide sample of deliveries. Anesth Analg 2010;110:1368–73. 3. Berg CJ, Callaghan WM, Syverson C, Henderson Z. Pregnancyrelated mortality in the United States, 1998 to 2005. Obstet Gynecol 2010;116:1302–9. 4. Japanese Ministry of Heath, Labour and Welfare. Change in maternal mortality rate in Japan. http://www.mhlw.go. jp/shingi/2008/11/dl/s1120-11n_0002.pdf#search [accessed June 2014]. 5. Mother’s & Children’s Health & Welfare Association. Maternal and child health statistics of Japan. Tokyo, Japan: Mother’s & Children’s Health Organization; 2011. 6. American College of Obstetricians and Gynecologists. ACOG Practice Bulletin, No. 76. Postpartum hemorrhage. Obstet Gynecol 2006;108:1039–47. 7. Charbit B, Mandelbrot L, Samain E, et al. The decrease of fibrinogen is an early predictor of the severity of postpartum hemorrhage. J Thromb Haemost 2007;5:266–73. 8. de Lloyd L, Bovington R, Kaye A, et al. Standard haemostatic tests following major obstetric haemorrhage. Int J Obstet Anesth 2011;20:135–41. 9. Gayat E, Resche-Rigon M, Rossignol M, et al. Predictive factors of advanced interventional procedures in a multicentre severe postpartum haemorrhage study. Intensive Care Med 2011;37:1816–25. 10. Cortet M, Deneux-Tharaux C, Dupont C, et al. Association between fibrinogen concentration and severity of postpartum haemorrhage: secondary analysis of a prospective trial. Br J Anaesth 2012;108:984–9. 11. Butwick AJ. Postpartum hemorrhage and low fibrinogen concentrations: the past, present and future. Int J Obstet Anesth 2013;22:87–91. 12. Simon L, Santi TM, Sacquin P, Hamza J. Pre-anaesthetic assessment of coagulation abnormalities in obstetric patients; usefulness, timing and clinical implications. Br J Anaesth 1997;78:678–83. 13. Guidelines for obstetrical practice in Japan: Japan Society of Obstetrics and Gynecology (JSOG) and Japan Association of Obstetricians and Gynecologists (JAOG) 2011 edition [in Japanese]. http://www.jsog.or.jp/activity/pdf/gl_sanka_2011.pdf [accessed March 2014]. 14. Ohkuchi A, Onagawa T, Usui R, et al. Effect of maternal age on blood loss during parturition: a retrospective multivariate analysis of 10,053 cases. J Perinat Med 2003;31:209–15. 15. Peyvandi F, Biguzzi E, Franchi F, et al. Elevated prepartum fibrinogen concentrations are not associated with a reduced risk of postpartum hemorrhage. J Thromb Haemost 2012;10: 1451–3. 16. Hui C, Lili M, Libin C, et al. Changes in coagulation and hemodynamics during pregnancy: a prospective longitudinal study of 58 cases. Arch Gynecol Obstet 2012;285:1231–6. 17. Hahn L, Korsan-Bengtsen K. The coagulation system during caesarean section. Acta Obstet Gynecol Scand 1975;54:49–55. 18. Bonnar J, Prentice CR, McNicol GP, Douglas AS. Haemostatic mechanism in the uterine circulation during placental separation. BMJ 1970;2:564–7. 19. Spahn DR, Cemy V, Coats TJ, et al. Management of bleeding following major trauma: a European guideline. Crit Care 2007;11:R17.
370 20. British Committee for Standards in Haematology, Stainsby D, MacLennan S, Thomas D, Issac J, Hamilton PJ. Guidelines on the management of massive blood loss. Br J Haematol 2006;135:634–41. 21. Bolliger D, Szlam F, Molinaro RJ, Rahe-Meyer N, Levy JH, Tanaka KA. Finding the optimal concentration range for
Antenatal fibrinogen and postpartum haemorrhage fibrinogen replacement after severe haemodilution: an in vitro model. Br J Anaesth 2009;102:793–9. 22. WHO guidelines for the management of postpartum haemorrhage and retained placenta. http://whqlibdoc.who.int/publications/ 2009/9789241598514_eng.pdf#search=’WHO+postpartum+hemorrhage+guideline’ [accessed March 2014].
ORIGINAL ARTICLE
www.obstetanesthesia.com
Antenatal fibrinogen concentrations and postpartum haemorrhage T. Yamada, R. Akaishi, Y. Oda, R. Nishida, T. Yamada, S. Ishikawa, M. Morikawa, T. Kojima, H. Minakami Department of Obstetrics, Hokkaido University Graduate School of Medicine, Sapporo, Hokkaido, Japan ABSTRACT Background: It is unclear whether antenatal fibrinogen concentrations are associated with postpartum haemorrhage. Methods: This retrospective study included 871 women with a singleton pregnancy but no known risk factors for postpartum haemorrhage, in whom fibrinogen concentration was measured within the 21 days before delivery. Correlation between antenatal fibrinogen concentrations and estimated blood loss was analysed. We tested the hypothesis that the risk of postpartum haemorrhage was higher in women with antenatal fibrinogen concentrations of <3.3 g/L. Postpartum haemorrhage was defined as an estimated blood loss P700 mL following vaginal delivery and P1000 mL following caesarean delivery. Results: In women delivering vaginally (n = 337), estimated blood loss tended to increase with decreasing antenatal fibrinogen concentration (R = 0.107, P = 0.05), median fibrinogen concentration was significantly lower in 69 women with postpartum haemorrhage than in 268 women without postpartum haemorrhage (3.93 vs. 4.18 g/L, P = 0.025), and postpartum haemorrhage occurred significantly more often in women with fibrinogen concentrations <3.3 g/L than in those with concentrations P3.3 g/L (38% [11/29] vs. 19% [58/308], P = 0.018). In women undergoing caesarean delivery (n = 534), median fibrinogen concentration did not differ between those who experienced postpartum haemorrhage (n = 128) and those who did not (n = 406) (4.18 g/L vs. 4.07 g/L, P = 0.43). Antenatal fibrinogen concentrations of <3.3 g/L were not associated with higher rates of postpartum haemorrhage (26% [11/43] vs. 24% [117/491], P = 0.80). Conclusions: Antenatal fibrinogen concentration <3.3 g/L may be a risk factor for postpartum haemorrhage among women following vaginal delivery. c 2014 Elsevier Ltd. All rights reserved.
Keywords: Postpartum haemorrhage; Fibrinogen; Vaginal delivery; Caesarean delivery
Introduction Postpartum haemorrhage (PPH) is the leading cause of maternal mortality, accounting for approximately one third of maternal deaths in Africa and Asia,1 and it remains a major cause of maternal mortality in developed countries. PPH has been associated with 12.5–19.1% of maternal mortality in the USA.2,3 In Japan, the maternal mortality rate is <5 per 100 000 births according to the Japanese Ministry of Heath, Labour and Welfare4 with PPH accounting for 14.1%, 9.7% and 6.7% of all causes of maternal mortality in Japan in 2000, 2005 and 2010, respectively.5 Although risk factors such as prolonged, augmented or rapid labour, history of PPH, preeclampsia, distended uterus and operative delivery are associated Accepted June 2014 Correspondence to: Takashi Yamada, Department of Obstetrics, Hokkaido University Graduate School of Medicine, Kita-ku N14 W6, Sapporo 060-8638, Japan. E-mail address: [email protected]
with PPH,6 it can occur without warning. Causes such as uterine rupture, uterine inversion, retained placenta and birth canal injury may be identified; however, the leading aetiology of PPH is uterine atony, and many women who develop PPH from uterine atony do not have known risk factors.2 There is interest in how the maternal coagulation profile alters during the course of PPH and fibrinogen has been identified an important factor that may influence the magnitude of blood loss.7–11 Studies have demonstrated that women who develop severe PPH have a low fibrinogen concentration in the initial stage of haemorrhage, that fibrinogen concentration in the initial stages of haemorrhage is a significant predictor of severity of PPH, and that decreases in fibrinogen concentration are closely correlated with the severity of blood loss. These studies have not determined whether a low fibrinogen concentration in the initial stages of PPH is caused by haemorrhage itself or reflects fibrinogen concentration before labour. It is unclear whether a low antenatal fibrinogen concentration is associated with PPH.
366
Antenatal fibrinogen and postpartum haemorrhage
This retrospective study was conducted to determine whether women with lower antenatal fibrinogen concentrations had a higher risk of developing PPH in the absence of known risk factors for PPH.
Methods This retrospective study was conducted with the approval of the Ethics Committee of Hokkaido University Hospital. The institutional delivery database was searched for women who gave birth to a singleton infant at or after gestational week 32 at Hokkaido University Hospital between 1 April 2007 and 31 March 2013. Women who had their fibrinogen levels checked in the 21 days pre-delivery and who had no known risk factors for PPH were included. Women were divided into two groups according to whether they had a vaginal or caesarean delivery. PPH was defined as an estimated blood loss (EBL) P700 mL for vaginal delivery and P1000 mL for caesarean delivery. Laboratory tests, including full blood count, biochemistry and coagulation-fibrinolysis are performed for the majority of pregnant women at our clinic and for those admitted to hospital for management of obstetric and incidental complications. All blood fibrinogen data from the institutional central laboratory were obtained for each patient during the study period. In some women, fibrinogen concentration was measured several times within 21 days of delivery. The fibrinogen value determined on the day closest to delivery was used. Demographic characteristics and EBL within 24 h of delivery were recorded from medical charts. Fibrinogen concentration was measured using the Clauss thrombin clotting time method in citrated blood samples and a haematology analyzer (CS2000i, Sysmex Co. Ltd., Kobe, Japan) after centrifugation. Intra- and inter-assay coefficients of variation were <6% and 10%, respectively. Normal fibrinogen concentration at our institution, not specific for pregnancy, is 2–4 g/L.
Statistical analysis Linear regression analysis was used to test the hypothesis that the risk of PPH was higher in women with antenatal fibrinogen concentration of <3.3 g/L than in those with antenatal fibrinogen concentration of P3.3 g/L. Data are presented as median [range] or mean ± standard deviation (SD). Statistical analyses were performed using the JMP10 c statistical software package (SAS, Cary, NC, USA). ANOVA and Tukey–Kramer HSD (honestly significant difference) tests were used for comparison of means. The Wilcoxon/Kruskal–Wallis method was used for comparison of medians. Fisher’s exact test was used for comparison of categorical variables. In all analyses, P < 0.05 was considered statistically significant.
Results A total of 1473 women were identified: 766 following vaginal delivery and 707 after caesarean delivery (Fig. 1). After matching fibrinogen data with medical charts, 925 women were found to have had blood fibrinogen concentrations measured in the 21 days before delivery. Fifty-four women with known risk factors for PPH before or after delivery were excluded: 52 with placenta praevia or low-lying placenta with or without placenta accreta, one who underwent radical hysterectomy after caesarean delivery for treatment of uterine cervical cancer, and one with uterine inversion immediately after delivery. The study therefore looked at data from 871 women, representing 59% of the 1473 women originally identified. Of the 871 women, 337 delivered vaginally and 534 underwent caesarean delivery (Table 1). For the 337 women delivering vaginally, EBL within 24 h after delivery was P700 mL in 67 (20%), >600 mL in 94 (28%) and >500 mL in 142 (42%). For caesarean deliveries, EBL within 24 h after delivery was P1000 mL in 128 of 534 women (24%). In women delivering vaginally, median [range] fibrinogen concentration of those who had a PPH was significantly lower than that of women who did not (3.93 [2.57–5.71] g/L vs. 4.18 [2.08–6.30] g/L, P = 0.025). For caesarean delivery, there was no difference in median [range] fibrinogen concentration for 128 women who had a PPH compared to the 406 who did not (4.18 [2.57–6.66] g/L vs. 4.07 [1.79–8.98] g/L, P = 0.43). The distribution of fibrinogen concentrations is shown in Fig. 2. Distribution did not differ between women in the vaginal and caesarean delivery groups: mean fibrinogen concentrations for women delivering vaginally or by caesarean were similar (4.13 ± 6.7 g/L vs. 4.17 ± 7.0 g/L, P > 0.05). Fibrinogen concentrations of <3.0 g/L and <3.3 g/L were detected in 23 (2.6%) and 72 (8.3%) of all patients, respectively. Estimated blood loss tended to increase with decreasing antenatal fibrinogen concentration in 337 women with vaginal deliveries (R = 0.1067, P = 0.05) (Fig. 3), but not in 534 women with caesarean deliveries (R = 0.048, P = 0.26). Among women delivering vaginally, the frequency of PPH was significantly higher in women with fibrinogen concentrations of <3.3 g/L than in those with fibrinogen concentrations of P3.3 g/L (Fig. 4). A fibrinogen cut-off value of 3.3 g/L yielded sensitivity of 16% (11/ 69), specificity of 93% (250/268), positive predictive value of 38% (11/29), and negative predictive value of 81% (250/308) for PPH. In women undergoing caesarean delivery, the frequency of PPH did not differ significantly for antenatal fibrinogen concentration above or below 3.3 g/L.
T. Yamada et al.
367 Number of women with singleton pregnancies who gave birth at ≥ 32 weeks of gestation during the study period
1473 women (766 vaginal and 707 caesarean deliveries)
Available fibrinogen level within 21 days pre-delivery?
No (n=548)
925 women (338 vaginal and 587 caesarean deliveries)
Risk factors for postpartum haemorrhage?
Yes (n=54)
871 women (337 vaginal and 534 caesarean deliveries) Fig. 1
Table 1
Flow diagram showing criteria for selection
Patient characteristics
Age (years) Nulliparous Body mass index (kg/m2) Gestation (weeks) Gestation <37 weeks Fibrinogen (g/L) Estimated blood loss P500 mL P600 mL P700 mL P1000 mL Birthweight (<2500 g) Birthweight <2500 g
Vaginal delivery (n = 337)
Caesarean delivery (n = 534)
P value
32 [18–44] 198 (58.8%) 20.1 [14.1–41.9] 38 [32–42] 52 (15.4%) 4.11 [2.08–6.3]
33 [18–48] 278 (52.1%) 21.2 [15.0–46.6] 37 [32–42] 88 (16.5%) 4.09 [1.79–8.98]
<0.0001 0.053 <0.0001 <0.0001 0.68 0.71
142 (42.1%) 96 (28.5%) 69 (20.5%) 28 (8.3%) 2925 [1260–4420] 69 (20.5%)
413 (78.3%) 345 (64.6%) 271 (50.7%) 128 (24%) 2815 [1102–5586] 127 (23.8%)
<0.0001 <0.0001 <0.0001 <0.0001 0.0028 0.25
Data are median [range] or number (%).
Discussion This study demonstrated that women with lower antenatal fibrinogen concentrations who gave birth vaginally were at higher risk of PPH; this was not seen in women undergoing caesarean delivery. To our knowledge, this has been noted in only one previous study.12 Japanese Guidelines for Obstetrical Practice13 report a 90th percentile value of EBL after vaginal delivery of 800 mL. In this study, EBL P500 mL, P600 mL and P700 mL occurred in 42%, 28% and 20% respectively, of the 337 women with vaginal delivery, findings consis-
tent with local recommendations. In addition, since only 0.3% of Japanese women require blood transfusion for excessive blood loss after vaginal delivery,14 it may be reasonable to define PPH following vaginal delivery as an EBL of P700 mL for Japanese women. Fibrinogen concentrations <3 g/L and <3.3 g/L were seen in 2.6% and 8.3% of all patients in the study. Simon et al. reported that 4.1% of 677 women delivering vaginally had an antenatal fibrinogen concentration <2.9 g/L. PPH, defined as clinically severe bleeding requiring volume expansion and manual uterine exploration, associated with either blood transfusion or a decrease in haemoglobin concentration P2.0 g/
Antenatal fibrinogen and postpartum haemorrhage Percentage (number) of women
368 35
(265)
30
(243)
25 20
(138) 15
(110)
10 5
(60) (32)
(23)
0 <3 -299
3-3.49 350-399 3.5-3.99 400-449 4-4.49 450-499 4.5-4.99500-549 5-5.49 550>5.5 300-349
Fibrinogen level (g/L)
Fig. 2 Fibrinogen concentrations in the 21 days before delivery
n=337
Estimated blood loss (mL)
R=−0.11 P=0.0504
2
3 4 5 6 Antenatal fibrinogen level (g/L)
Fig. 3 Correlation between antenatal fibrinogen concentrations and estimated blood loss
P=0.018
Percentage of women with postpartum haemorrhage
40
(11/29)
30 20
(11/43)
(117/491)
(58/308)
10 0
<3.3g/L ≥ 3.3g/L Vaginal delivery
<3.3g/L ≥ 3.3g/L Caesarean delivery
Fig. 4 Frequency of postpartum haemorrhage in vaginal and caesarean delivery groups analysed using an antenatal fibrinogen cut-off of 3.3 g/L
dL, or both, within 24 h of delivery occurred in 6.8% of the 677 women. A cut-off fibrinogen value of 2.9 g/L yielded a sensitivity of 19.6% and a positive-predictive value of 39.1% for PPH.12 Using our criteria for PPH of an EBL P700 mL, the cut-off of 3.3 g/L yielded sensitivity of 16% and positive-predictive value of 38% for prediction of PPH following vaginal delivery. Peyvandi et al. examined the risk of PPH according to increasing antenatal fibrinogen concentration in 4461 women and failed to show any difference in five groups (fibrinogen concentration 64 g/L, 4.01–4.5 g/L, 4.51–5 g/L, 5.01–5.50 g/L and >5.51 g/L).15 In the present study, the risk of PPH did not differ significantly between women with fibrinogen levels above or below 3.6 g/L (data not shown), consistent with the results reported by Peyvandi et al. However, an antenatal fibrinogen concentration of 4 g/L was the 20th percentile value in Peyvandi’s study, while in our study it corresponded to the 43rd percentile value, and our 20th percentile fibrinogen value was 3.6 g/L. The reason for this difference is not clear. Fibrinogen concentration increases during pregnancy from a non-pregnant concentration of 2.52 ± 0.7 g/L to 4.47 ± 0.9 g/L by 33 weeks of gestation. 16 Fibrinogen is consumed at the time of caesarean delivery. Uterine vein fibrinogen concentration immediately before uterine wall incision falls from 4.25 ± 0.21 to 4.1 ± 0.2 g/L and 4.05 ± 0.22 g/L, 5 and 15 min after removal of the placenta. Fibrinogen concentration in the uterine vein after removal of the placenta is significantly lower than that in the circulating blood,17 and its concentration in circulating blood decreases from approximately 4.0 g/L to 3.46 g/L within 4 h of caesarean delivery.18 Current guidelines recommend correcting fibrinogen concentration if <1.0 g/L,19,20 although the rate of clot formation is not optimised until fibrinogen concentrations reach >2.0 g/L in an in vitro haemodilution model.21 Studies have suggested that the fibrinogen concentration at an early stage of PPH is associated with severity of haemorrhage,7,10 and a lower fibrinogen concentration is likely to occur with increasing blood loss.9 The risk of severe PPH has been shown to be 2.63-fold higher for each 1 g/L decrease in fibrinogen concentration at the beginning of PPH.7 Mean fibrinogen concentration at diagnosis of PPH has been shown to be significantly lower in women with subsequent worsening of PPH compared to that in women without worsening of PPH (3.4 ± 0.9 g/L vs. 4.2 ± 1.2 g/L).10 The lowest fibrinogen concentration measured within 24 h of delivery has the best correlation with increasing severity of PPH.8 In one study, median fibrinogen concentration on admission for treatment of PPH was significantly lower in women requiring intervention, such as uterine artery embolisation and/or haemostatic surgery, than in those who did not undergo such procedures (1.8 g/L vs. 2.5 g/L, P < 0.0001).9
T. Yamada et al. Women undergoing caesarean delivery who had low antenatal fibrinogen concentration did not have a higher risk of PPH in this study. Uterine atony is a leading cause of PPH after vaginal delivery.2,22 Although uterine atony is ill-defined, insufficient uterine involution is thought to lead to PPH. Uterine atony is responsible for 79% of PPH events in the USA.2 Continuous uterine massage facilitates involution and is recommended for the treatment of PPH, but not as an intervention to prevent PPH after vaginal delivery in women who have received prophylactic oxytocin because uterine massage may cause maternal discomfort.22 Our routine procedures include continuous uterine massage during caesarean delivery, but not after vaginal delivery. This may have reduced the effect on PPH of low antenatal fibrinogen concentration following caesarean delivery in the present study. The study has some limitations. Selection bias is a potential concern. There were 766 vaginal deliveries during the study period, but only 337 women (44%) were included in the study; this was mainly due to limited or absent fibrinogen data within the 21 days before delivery. This suggests that the final three weeks of pregnancy were uneventful in 56% of women excluded from the present analysis. However, our protocol for management of pregnant women did not include treatment options that were determined by antenatal fibrinogen concentration. It seems reasonable to speculate that obstetric management did not differ from those women in whom the fibrinogen concentration was unknown. Many clinical factors may affect blood loss after vaginal delivery, and it was not clear from this study whether a low fibrinogen concentration was an independent risk factor for PPH. In conclusion, the present study suggests that women who deliver vaginally with a relatively low antenatal fibrinogen concentration had a higher risk of developing PPH. However, it remains unclear whether antenatal low fibrinogen concentration was an independent risk factor for PPH. Larger prospective studies are needed to clarify this issue.
Disclosure This study was supported by departmental funds only and the authors have no conflicts of interest to declare.
Acknowledgement We thank the staff of the Department of Obstetrics, Hokkaido University Hospital, for their cooperation in blood sampling.
369
References 1. Khan KS, Wojdyla D, Say L, Gu¨lmezoglu AM, Van Look PF. WHO analysis of causes of maternal death: a systematic review. Lancet 2006;367:1066–74. 2. Bateman BT, Berman MF, Riley LE, Leffert LR. The epidemiology of postpartum hemorrhage in a large, nationwide sample of deliveries. Anesth Analg 2010;110:1368–73. 3. Berg CJ, Callaghan WM, Syverson C, Henderson Z. Pregnancyrelated mortality in the United States, 1998 to 2005. Obstet Gynecol 2010;116:1302–9. 4. Japanese Ministry of Heath, Labour and Welfare. Change in maternal mortality rate in Japan. http://www.mhlw.go. jp/shingi/2008/11/dl/s1120-11n_0002.pdf#search [accessed June 2014]. 5. Mother’s & Children’s Health & Welfare Association. Maternal and child health statistics of Japan. Tokyo, Japan: Mother’s & Children’s Health Organization; 2011. 6. American College of Obstetricians and Gynecologists. ACOG Practice Bulletin, No. 76. Postpartum hemorrhage. Obstet Gynecol 2006;108:1039–47. 7. Charbit B, Mandelbrot L, Samain E, et al. The decrease of fibrinogen is an early predictor of the severity of postpartum hemorrhage. J Thromb Haemost 2007;5:266–73. 8. de Lloyd L, Bovington R, Kaye A, et al. Standard haemostatic tests following major obstetric haemorrhage. Int J Obstet Anesth 2011;20:135–41. 9. Gayat E, Resche-Rigon M, Rossignol M, et al. Predictive factors of advanced interventional procedures in a multicentre severe postpartum haemorrhage study. Intensive Care Med 2011;37:1816–25. 10. Cortet M, Deneux-Tharaux C, Dupont C, et al. Association between fibrinogen concentration and severity of postpartum haemorrhage: secondary analysis of a prospective trial. Br J Anaesth 2012;108:984–9. 11. Butwick AJ. Postpartum hemorrhage and low fibrinogen concentrations: the past, present and future. Int J Obstet Anesth 2013;22:87–91. 12. Simon L, Santi TM, Sacquin P, Hamza J. Pre-anaesthetic assessment of coagulation abnormalities in obstetric patients; usefulness, timing and clinical implications. Br J Anaesth 1997;78:678–83. 13. Guidelines for obstetrical practice in Japan: Japan Society of Obstetrics and Gynecology (JSOG) and Japan Association of Obstetricians and Gynecologists (JAOG) 2011 edition [in Japanese]. http://www.jsog.or.jp/activity/pdf/gl_sanka_2011.pdf [accessed March 2014]. 14. Ohkuchi A, Onagawa T, Usui R, et al. Effect of maternal age on blood loss during parturition: a retrospective multivariate analysis of 10,053 cases. J Perinat Med 2003;31:209–15. 15. Peyvandi F, Biguzzi E, Franchi F, et al. Elevated prepartum fibrinogen concentrations are not associated with a reduced risk of postpartum hemorrhage. J Thromb Haemost 2012;10: 1451–3. 16. Hui C, Lili M, Libin C, et al. Changes in coagulation and hemodynamics during pregnancy: a prospective longitudinal study of 58 cases. Arch Gynecol Obstet 2012;285:1231–6. 17. Hahn L, Korsan-Bengtsen K. The coagulation system during caesarean section. Acta Obstet Gynecol Scand 1975;54:49–55. 18. Bonnar J, Prentice CR, McNicol GP, Douglas AS. Haemostatic mechanism in the uterine circulation during placental separation. BMJ 1970;2:564–7. 19. Spahn DR, Cemy V, Coats TJ, et al. Management of bleeding following major trauma: a European guideline. Crit Care 2007;11:R17.
370 20. British Committee for Standards in Haematology, Stainsby D, MacLennan S, Thomas D, Issac J, Hamilton PJ. Guidelines on the management of massive blood loss. Br J Haematol 2006;135:634–41. 21. Bolliger D, Szlam F, Molinaro RJ, Rahe-Meyer N, Levy JH, Tanaka KA. Finding the optimal concentration range for
Antenatal fibrinogen and postpartum haemorrhage fibrinogen replacement after severe haemodilution: an in vitro model. Br J Anaesth 2009;102:793–9. 22. WHO guidelines for the management of postpartum haemorrhage and retained placenta. http://whqlibdoc.who.int/publications/ 2009/9789241598514_eng.pdf#search=’WHO+postpartum+hemorrhage+guideline’ [accessed March 2014].