2018 Volume 6 Issue 1 Pages 20-25
Aim: This study aimed to identify risk factors for bleeding complications of postoperative prophylactic anticoagulation after cesarean section in preeclampsia cases.
Methods: A total of 68 cases of preeclampsia or superimposed preeclampsia at a tertiary perinatal center in Tokyo between 2012 and 2017 were recruited for this study. Bleeding complications were defined as subcutaneous, subfascial, or intraperitoneal hematoma detected by ultrasonography or computed tomography. Associations of clinical and laboratory data with bleeding complications were assessed by univariate and multivariate analyses.
Results: Bleeding complications were recorded in nine cases: subcutaneous hematoma in four cases, subfascial hematoma in four cases, and intraperitoneal hematoma in one case. Univariate analysis revealed preoperative platelet count and 24-h urine protein level to be associated with bleeding complications. Moreover, multivariate logistic regression analysis revealed preoperative platelet count (odds ratio, 0.867; 95% confidential interval, 0.756–0.994; P=0.04) and 24-h urine protein level (odds ratio, 1.498; 95% confidential interval, 1.031–2.176; P=0.03) to be independent risk factors for bleeding complications.
Conclusion: Preoperative platelet count and 24-h urine protein level may help to identify patients at increased risk for bleeding complications.
Venous thromboembolism (VTE), which includes deep vein thrombosis (DVT) and pulmonary embolism (PE), is a leading cause of maternal morbidity and mortality in Japan.1) Not only pregnancy itself, but other factors during pregnancy and postpartum increase the risk of VTE. Among these, cesarean section is one of the most important risk factors.2) In Japan, the incidences of DVT and PE after cesarean section are 0.04% and 0.06%, respectively, and cesarean section increases the risk of developing DVT and PE by 5- and 22-fold, respectively.2,3) As cesarean section has become more common in Japan, preventing VTE after the procedure is becoming all the more important.
Preeclampsia, a clinical entity included in hypertensive disorders of pregnancy (HDP), is characterized by hypertension, proteinuria, and multiple organ complications. Although the exact pathogenesis of preeclampsia is still unknown, it is thought to involve endothelial dysfunction resulting from an imbalance of pro-angiogenic and anti-angiogenic factors.4) As maternal endothelial damage activates the coagulation system, cesarean section in preeclampsia may increase the risk of VTE.5) Consideration of prophylactic anticoagulation therapy for women with preeclampsia after cesarean section is recommended by the guidelines of the Japan Society of Obstetrics and Gynecology (JSOG) and Japan Association of Obstetricians and Gynecologists (JAOG).6)
While preeclampsia causes a hypercoagulable state, it is often complicated by liver dysfunction, renal dysfunction, uncontrolled hypertension, and disseminated intravascular coagulation (DIC).4) These complications may increase the risk of bleeding events associated with anticoagulation therapy. In a previous survey conducted in 66 hospitals in Japan, 9% of hospitals reported accidental complications associated with anticoagulation therapy.7)
In order to reduce the risk of bleeding events, the Japan Society for the Study of Hypertension in Pregnancy (JSSHP) recommends evaluating body mass index (BMI), renal function, and blood pressure of women with HDP before prophylactic anticoagulation therapy.8) However, there is insufficient clinical evidence on preventive anticoagulation therapy for preeclamptic women who are at higher risk for bleeding complications. Thus, this study assessed risk factors for bleeding complications of prophylactic anticoagulation therapy after cesarean section in preeclampsia cases.
A retrospective study was conducted at Tokyo Metropolitan Bokutoh Hospital, a tertiary perinatal center in Tokyo. The Diagnosis Procedure Combination (DPC) database was used to identify cases of HDP in which cesarean section was performed between June 2012 and May 2017. Medical records were reviewed to confirm whether the diagnosis of preeclampsia or superimposed preeclampsia was correct according to JSSHP criteria.8) Cases of fetal growth restriction without any other maternal features of preeclampsia were excluded. This study was approved by the institutional review board of Tokyo Metropolitan Bokutoh Hospital.
Pharmacological and non-pharmacological methods were used to prevent VTE. Compression stockings were worn during the perioperative period. Intermittent pneumatic compression devices were set postoperatively and removed by postoperative day 1. As a prophylactic anticoagulant, 5,000 units of unfractionated heparin (UFH) was administered subcutaneously; once six h after the operation and/or at 6 a.m. on postoperative day 1, and every 12 h afterwards for two days. UFH was discontinued when bleeding complications were observed during anticoagulation therapy.
All cases were divided into two groups based on the presence or absence of postoperative bleeding complications. Bleeding complications were defined as subcutaneous, subfascial, and intraperitoneal hematoma detected by ultrasonography or computed tomography. It was left to the doctors’ discretion whether to perform imaging examinations when signs and symptoms suggesting bleeding complications were noted. The following factors were analyzed and compared between the two groups: age, preoperative BMI, use of intravenous antihypertensive agents, intraoperative blood loss, symptomatic VTE, and preoperative laboratory data such as serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine, platelet count, activated partial thromboplastin time (APTT), prothrombin time (PT) activity, fibrinogen, 24-h urine protein level, 24-h urine volume, and creatinine clearance.
Exclusion criteriaThe following cases were excluded from this study: 1) lack of data on factors of interest; 2) lack of postoperative UFH; 3) discontinuation, dose reduction, or extended-interval dosing of UFH due to reasons other than bleeding complications during prophylactic anticoagulation therapy.
Statistical analysisContinuous variables are presented as median and interquartile range (IQR), and categorical variables as number and percent. Continuous variables were analyzed with the Mann-Whitney U-test, and categorical variables were analyzed with Fisher’s exact test. Variables with statistical differences were selected to perform multivariate logistic regression analysis for independent risk factors. Box-Cox transformations were performed on the selected variables to account for non-normality. P < 0.05 was considered statistically significant. All statistical analyses were performed using JMP 14 (SAS Institute Inc., Cary, NC, USA).
Preeclampsia and superimposed preeclampsia accounted for 6.1% (94/1,536) of the total cesarean sections performed in the 5-year study period. Prophylactic anticoagulation therapy was not performed in one case due to decreased renal function. Postoperative bleeding complications were recorded in 11 of 93 cases (11.8%). Among the 93 cases, data were lacking in 21. The discontinuation, dose reduction, or extended-interval dosing of heparin was recorded in four cases. The final study population consisted of nine cases in the bleeding group and 59 in the non-bleeding group (Figure 1).
Selection of study population.
Cesarean sections were performed in 1,536 cases during the 5-year study period. Prophylactic anticoagulation therapy was performed in 93 cases of preeclampsia or superimposed preeclampsia. The final study population consisted of nine cases in the bleeding group and 59 cases in the non-bleeding group.
Bleeding complications included four subcutaneous hematomas, four subfascial hematomas, and one intraperitoneal hematoma. These complications were detected within four days after cesarean section. Reoperation to remove the hematoma was performed in two cases of subfascial hematoma. An interventional radiology procedure was performed to arrest bleeding from the left superior vesical artery in one case of intraperitoneal hematoma.
Univariate analysis revealed that preoperative platelet counts were significantly lower in the bleeding group (P=0.03) and that levels of preoperative 24-h urine protein were significantly higher in the bleeding group (P=0.04), as compared to the non-bleeding group (Table 1A). Although cases in the bleeding group tended to have low BMIs and high serum levels of AST and creatinine, there were no significant differences in these factors between the two groups. In multivariate logistic regression analysis, both preoperative platelet count (odds ratio, 0.867; 95% confidential interval, 0.756–0.994; P=0.04) and 24-h urine protein level (odds ratio, 1.498; 95% confidential interval, 1.031–2.176; P=0.03) were independent risk factors for bleeding complications.
| Bleeding group (n=9) | Non-bleeding group (n=59) | P value | |||
|---|---|---|---|---|---|
| Age (years) | 36 | (34.0–37) | 36 | (33.5–38) | 0.70 |
| BMI (kg/m2) | 23.5 | (23.00–25.2) | 26.8 | (24.05–29.0) | 0.06 |
| Use of intravenous antihypertensive agents (n) | 3 | (33.3%) | 24 | (40.7%) | 1.00 |
| Intraoperative blood loss (ml) | 475 | (385–657) | 700 | (437–905) | 0.17 |
| Symptomatic VTE (n) | 0 | (0%) | 0 | (0%) | 1.00 |
| ALT (IU/l) | 27 | (14.0–40.0) | 15 | (9.5–24.5) | 0.07 |
| AST (IU/l) | 30 | (19–40.0) | 21 | (16–28.5) | 0.10 |
| Cre (mg/dl) | 0.8 | (0.7–0.9) | 0.7 | (0.6–0.8) | 0.09 |
| Plt (104/μl) | 16.2 | (12.5–16.7) | 19.3 | (16.0–24.0) | 0.03 |
| APTT (sec) | 32.9 | (27.5–33.3) | 28.7 | (27.0–30.1) | 0.13 |
| PT activity below the lower limit of normal (<75%) (n) | 0 | (0%) | 0 | (0%) | 1.00 |
| Fib (mg/dl) | 388 | (343.0–440.0) | 410 | (337.5–473.5) | 0.62 |
| 24-h urine protein (g) | 5.248 | (3.339–6.719) | 2.373 | (0.748–5.591) | 0.04 |
| 24-h urine volume (ml) | 1,850 | (1,600–2,800) | 1,600 | (1,200–2,100) | 0.37 |
| Cre clearance (ml/min) | 89.2 | (80.7–93.1) | 100.9 | (78.0–126.7) | 0.30 |
| Odds ratio | 95% CI | P value | |
|---|---|---|---|
| Plt (104/μl) | 0.867 | 0.756–0.994 | 0.04 |
| 24-h urine protein excretion (g) | 1.498 | 1.031–2.176 | 0.03 |
Continuous variables are presented as median and interquartile range, and categorical variables as number and percent. Box-Cox transformations were performed on variables (Plt and 24-h urine protein excretion level) for the multivariate analysis.
BMI, body mass index; VTE, venous thromboembolism; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Cre, creatinine; Plt, platelet; APTT, activated partial thromboplastin time; PT, prothrombin time; Fib, fibrinogen; CI, confidence interval.
The incidence of bleeding complications in cases of preeclampsia and superimposed preeclampsia treated with postoperative prophylactic anticoagulation therapy was 11.8% (11/93) in this study. Only a few reports have been published on bleeding complications during postoperative anticoagulation therapy that focused on preeclamptic women7). The incidence of bleeding events has been reported to range from 0.7–1.4%9,10) in preeclamptic cases administered postoperative prophylactic UFH. The following reasons may account for the discrepancy in incidence of bleeding complications between previous studies and the present study. First, the definition of bleeding complications in the present study differed from those of the previous reports. Because it was left to the doctors’ discretion whether imaging examinations were performed and the hematomas detected by imaging examinations were regarded as bleeding complications regardless of size, the present study may have overestimated the incidence of bleeding complications. Second, the previous reports did not limit their subjects to preeclamptic women; all women after cesarean section received heparin, excluding those at risk of developing bleeding complications. On the other hand, the present study limited subjects to preeclamptic women and found a higher incidence of bleeding complications. This may suggest that women with preeclampsia undergoing cesarean section are at increased risk of developing bleeding complications after prophylactic anticoagulation therapy.
Conservative treatment may be reasonable for cases of subcutaneous hematoma, but some cases of active bleeding, severe pain, or severe anemia may require invasive treatment such as reoperation or interventional radiology procedures. Intra-abdominal bleeding and hematoma have been reported as the leading indication for reoperation after cesarean section.11) However, it is sometimes difficult to detect the bleeding point in reoperation due to poor visualization of the surgical field.12) In the present study, reoperation was performed to remove hematomas and to relieve pain arising from compression by hematomas, rather than to achieve hemostasis. Currently, interventional radiology is preferred as a minimally invasive treatment and due to its high success rate in achieving hemostasis.12) In this study, interventional radiology was used to stop bleeding from the left superior vesical artery in one case of intraperitoneal hematoma.
The relationship between proteinuria and adverse maternal outcomes in preeclampsia has been reported in many studies.13,14,15,16) Although some of these studies reported that heavy proteinuria increases adverse maternal outcomes including reoperation, blood transfusion, acute renal failure, and thrombocytopenia,13,14) others suggest that the extent of proteinuria is not associated with adverse maternal outcomes.15,16) The present study found 24-h urine protein level to be an independent risk factor for bleeding events. The extent of proteinuria, which may reflect renal endothelial damage, may also reflect vascular vulnerability, which can lead to bleeding events.
Preeclamptic women tend to have lower platelet counts than normal pregnant women.17) A previous report suggested that increased platelet turnover caused by endothelial injury may be one of the reasons for the lower platelet count.17) Once preeclampsia is complicated by HELLP syndrome, consumption of platelets may increase and promote hemorrhagic tendency. Although thrombocytopenia is defined as a platelet count less than 100,000/μl according to the diagnostic standard proposed by Sibai et al.,18) the results of the present study suggest that preeclamptic women should be recognized as being at high risk for bleeding complications by anticoagulant therapy even when preoperative platelet counts are not abnormally low.
According to Japanese drug manufacturers, heparin should not be administered or should be administered with caution to patients with severe hypertension because heparin may cause injured vessels to bleed. The JSSHP also recommends that prophylactic anticoagulants should not be used or postponed in women with poorly controlled hypertension.8) In the present study, a significant difference was not observed in the use of intravenous antihypertensive agents, which was used as a surrogate variable for severe hypertension because antihypertensive agents were intravenously administered in all severe hypertension cases. Since blood pressure was under control in all cases after the initiation of intravenous injection therapy, the use of intravenous antihypertensive agents may not have reflected uncontrolled hypertension.
Cases of preeclampsia which required discontinuation, dose reduction, or extended-interval dosing of heparin for reasons other than bleeding were excluded from this study. Most of these cases had reduced renal function, which was the reason for their exclusion. This may explain the lack of a significant difference in renal function between the bleeding and non-bleeding groups. Because the clearance of heparin from blood could be delayed in patients with renal impairment,19) heparin should be administered carefully to such patients. However, there have been no guidelines or proposals on optimizing heparin administration for preeclamptic women with renal impairment. A survey involving 66 hospitals revealed that prophylactic anticoagulation therapy was modified and optimized for women with HDP in only half of the hospitals.7) In the present study, the optimization of heparin administration for cases of renal impairment was left to the attending physicians’ discretion.
One of the limitations of this study is its retrospective design. Moreover, the exclusion of many cases decreased the statistical power of the analysis. Excluded cases included 21 cases with missing data, particularly with respect to the 24-h urine collection test. The spot urinary protein/creatinine ratio (P/C) is currently used widely as a rapid alternative test to the 24-h urine collection test.20) Thus, further studies using the P/C ratio may be helpful for evaluating the relationship between proteinuria and bleeding complications in emergent cases. Second, the number of bleeding complications may have been underestimated since whether or not to perform imaging examinations was left to the doctors’ discretion. The third limitation relates to the dosage and administration of prophylactic anticoagulants. UFH was administered as an anticoagulant, but low-molecular-weight heparins (LMWH) are now widely used as an alternative for prophylactic anticoagulation therapy.21) The advantages of LMWH over UFH include lower risks of bleeding complications, heparin-induced thrombocytopenia, and osteoporosis.22) Although the Royal College of Obstetricians and Gynaecologists (RCOG) reported contraindications for LMWH use as an anticoagulant, this recommendation is based on evidence from non-pregnant populations.21) Further studies may be needed to assess the impact of anticoagulants other than UFH on bleeding complications in preeclampsia.
In conclusion, preoperative platelet count and degree of proteinuria should be considered when prophylactic anticoagulation therapy is performed in women with preeclampsia after cesarean section. Even after initiating anticoagulation therapy, patients should be followed carefully with physical examinations, blood tests, and ultrasounds for the early detection of bleeding complications.
None.
The authors hereby declare that there are no conflicts of interest regarding the contents of this article.
