Pharmacologic thromboprophylaxis following cesarean delivery-what is the evidence? A critical reappraisal

Introduction

Cesarean delivery rates have steadily risen over the past decade [1]. In Germany, more than 30 % of pregnant women were delivered by c.d. in 2023 [2].

Venous thromboembolism (VTE) is a leading cause of maternal mortality accounting for 15 % of maternal deaths in developed countries [3]. More than 80 % of maternal deaths due to VTE are related to c.d. [4]. Pregnancy and delivery fulfil the criteria of.

Virchow`s triad – hypercoagulability, venous stasis and vascular damage associated with a 15–35-fold increase in the risk of VTE compared with non-pregnant, age-matched women [5].

Cesarean delivery further increases the risk of VTE compared with uncomplicated vaginal delivery due to a more extensive tissue damage associated with a more pronounced release of tissue factor leading to increased thrombin generation and clotting. Surgery-related immobilization associated with venous stasis is an additional risk factor for VTE in the postpartum period.

Notably, the incidence of VTE after c.d. is increased for more than 12 weeks after delivery [6], [7], [8]. Sultan et al. [9] reported on an incidence of 637 episodes per 100,000 person-years after c.d.; this corresponds to 147 VTE events up to 12 weeks postpartum.

Limitations of results from database regarding the increased risk of VTE after c.d. are lack of information of confounding factors and additional VTE risk factors [3].

According to a meta-analysis controlling for confounding factors the risk of VTE was found to be 4-fold higher after c.d. than after vaginal delivery and was considered as independent of other VTE risk factors [10]. The pooled incidences of VTE were 2.6–4.3/1,000 c.d. and the adjusted pooled odds ratio associated with c.d. was 2.5 (95 % CI 1.8–3.1). The pooled incidence was 1.6/1,000 pregnancies (95 % CI 1.2–2.2) for elective c.d. and 2.4/1,000 pregnancies (95 % CI 0.8–4.5) for emergency c.d. and the pooled odds ratios were 2.3 (95 % CI 1.7–3.1) and 3.6 (95 % CI 2.8–4.7, respectively, when compared to vaginal delivery. In women without pregnancy complications or other risk factors for VTE (n=1,336) the pooled incidence was 2.0 (95 % CI 0.2–5.2) which is consistent with the results of a recent meta-analysis (OR 2.05; 95 % CI 1.7–2.5) [11] and others [12], [13], [14] The Nationwide Readmissions Database reported on an incidence of 18 VTE per 1,000 c.d. among 2 347,711 deliveries [15].

Prophylactic anticoagulation treatment after c.d. represents a daily question for many clinicians [16]. Surveys from many countries have shown that post – cesarean pharmacologic thromboprophylaxis is a common practice [17], [18], [19].

However, benefit and harm of pharmacologic thromboprophylaxis after c.d. remain controversial. Hence, we evaluated the evidence of thromboprophylaxis efficacy in patients with c.d. and to comment on current guideline recommendations.

Search strategy

The search was performed in PubMed, Embase, Cochrane Library and Google Scholar between 2000 and 12/2024. Only article published in English were considered -The search keywords were as follows: venous thromboembolism, cesarean section/delivery, thromboprophylaxis, unfractionated heparin, low molecular weight heparin.

Comparison of LMWH with no treatment or placebo

Only four randomized controlled trials (RCT) and one prospective cohort study including a total of 1,196 patients assessed the thromboprophylaxis efficacy of LMWH vs. no treatment or placebo.

In a randomized controlled double-blind trial including 76 patients with an emergency or elective c.d. 39 women were given 2,500 IU dalteparin subcutaneously 4–24 h after c.d. for 4–5 days post-operatively and 37 patient received saline solution. Burrows et al. [20]; 2 and 6 weeks after c.d. the women were contacted by phone. One case of left iliofemoral thrombosis occurred after placebo and no VTE event after dalteparin prophylaxis.

The study has several limitations: the low recruitment rate of 20 % and the small sample size, the relatively low dose of dalteparin administered, no differentiation between emergency and elective c.d. and the low validity of VTE diagnosis after discharge from hospital.

A further prospective randomized study from Saudi Arabia included 300 patients with low-risk c.d., who received either 4,500 IU tinzaparin subcutaneously once daily 12–24 h after c.d. for 2 weeks post-operatively (n=100) or placebo (n=200). One case of deep vein thrombosis occurred with placebo and no VTE event was observed in the tinzaparin group. The study was published as an abstract without sufficient information on patient`s characteristics, additional risk factors for VTE, and the type (emergency or elective) of c.d. [21].

In a multicenter randomized controlled trial from UK including 141 patients with c.d. thromboprophylaxis enoxaparin (40 mg/d) administered approximately 6.5 h after c.d. and continued up to 14 days post-operatively (n=70) was compared to saline solution (n=71) [22]; 55 women in each group had c.d. performed while in active labor. The primary outcome criteria were symptomatic, confirmed VTE during the first 6 months after delivery and serious wound complications.

There was one pulmonary embolism confirmed by ventilation/perfusion scan two weeks after delivery in women treated with enoxaparin and one woman in each group had serious wound complications (wound infections). The authors emphasized that recruiting sufficient patients for randomized, placebo-controlled trials regarding thromboprophylaxis following c.d. may be difficult.

To confirm the thromboprophylaxis efficacy of LMWH the authors stated that up to 20,000 women need to be recruited assuming an incidence of 0.5 % in the placebo group and a relative risk of 0.5.

Limitations of the study are, in particular, the inadequate statistical power due to small number of patients included and the lack of information on follow-up of patients after discharge from hospital.

The main objective of a randomized trial (n=159) from Germany was to evaluate rheological changes after elective c.d. in association with the use of LMWH or unfractionated heparin [23]; 100 women received 5,000 IU dalteparin daily or 2 × 5,000 IU calciparin daily beginning 6 h after c.d. and continued for 7 days after delivery; 59 patients had no heparin. All patients were screened for VTE until discharge from hospital by impedance plethysmography or Doppler ultrasound. There was one case of deep vein thrombosis in the calciparin and one in the control group.

The study has several limitations: the primary outcome was not the incidence of VTE after c.d. but the influence of LMWH/unfractionated heparin on rheological properties of blood. Moreover, there are limited information on patient´s characteristics, additional risk factor for VTE and follow-up of patients after discharge from hospital.

An observational-longitudinal cohort study from Italy included a total of 529 patients with elective c.d. [24]. The women were 35–44 years old and had no significant other risk factors for VTE.

Either enoxaparin (40 mg/d) or dalteparin (5,000 IU/d) were administered 12 h after c.d. for 7 days post-operatively (n=349) and compared to women without LMWH thromboprophylaxis (n=180).

Mechanical prophylaxis (elastic stockings or intermittent pneumatic compression) was performed in all of patients.

The aim of the study was to evaluate the rate of thromboembolic events in both groups during hospitalization until 6 weeks after delivery.

Irrespective of the advanced maternal age, no thromboembolic events were found during the observation period. Maternal hemorrhage requiring transfusion was 4.7 % in the heparin groups and 0 in the control group (p=0.03). The authors concluded that maternal age is not a decision-making factor for or against LMWH prophylaxis after elective c.d. in an unscreened population.

Limitations of the study are the study design (not randomized), the lower number of patients in the control group and limited information on patient’s characteristics and follow-up period.

These five studies were evaluated in the meta-analysis by Yang et al. 2018 (Table 1). A Cochrane Systematic Review 2021 [25] included 5 randomized trials (1981–2015) with 1,147 women. No significant differences were found in the reduction of symptomatic thromboembolism events following c.d. between thromboprophylaxis using LMWH/unfractionated heparin compared to no treatment or placebo (RR 1.30; 95 % CI 0.39–4.27).

Comparison of different LMWHs, different doses and dosing intervals

There exist only four studies on post – cesarean thromboprophylaxis, which were not primarily focused on determining the incidence of VTE but evaluated other issues associated with heparin prophylaxis after c.d. The purpose of a randomized single-blind study was to measure factor Xa activity, thrombin-antithrombin complex and tissue factor pathway inhibitor in 30 women either receiving 5,000 IU dalteparin/day (n=10), enoxaparin 40 mg/day (n=10) or tinzaparin 50 IU/kg 6 h after c.d. for 5 days [26]. As was expected due to the very small sample size there were no thromboembolic events in all groups. LMWHs differed in their effect on haemostatic parameters.

In a randomized study thromboprophylaxis efficacy of 3,500 IU bemiparin once daily for 5 days (n=311) was compared to bemiparin 3,500 IU once daily for 10 days (n=335) following c.d. [27], 73.6 % of women had emergency c.d. Pulmonary embolism occurred in one patient in the 10-day regimen group (0.3 %).

A recent randomized controlled trial compared anti-Xa levels 4–6 h after at least the third anticoagulant dose in 146 women receiving either a fixed dose enoxaparin (40 mg daily for BMI <40 kg/m² or 40 mg every 12 h for BMI ≥40 kg/m²) or weight-based (0.5 mg/kg every 12 h) enoxaparin prophylaxis after c.d. [28]. No thromboembolic events were observed.

The weight-based enoxaparin regimen was superior to the fixed enoxaparin dosing in achieving prophylactic peak ant-Xa levels. These results are in line with other studies [29], 30] comparing the different dose regimens in obese women with c.d.

Thromboprophylaxis following c.d. according to decision and risk score models

Numerous decision-analytic and risk score models for thromboprophylaxis during pregnancy and the puerperium have been proposed weighing the benefits and harms of pharmacological prophylaxis, as recently systematically reviewed by Davis et al. [31].

However, there are only few publications on risk assessment models for VTE related to c.d. Blondon performed a decision analysis comparing one-week LMWH prophylaxis after c.d. to no prophylaxis in women with and without additional risk factors. He concluded that LMWH was only slightly more effective in women without risk factors, but in women with additional risk factors, reduction of VTE greatly outnumbered the increase in major hemorrhage [16].

The efficacy of a risk score model for preventing c.-d.-related VTE was evaluated in a recent prospective pilot study [32] The risk score included maternal age, weight, history of VTE, thrombophilia, immobility, parity and varicose vein; 233 (46.5 %) women were classified as low risk and had only mechanical prophylaxis, 268 (53.5 %) women at moderate-high risk were treated with enoxaparin (20–40 mg/d), starting 6–8 h after c.d. Women were followed-up for 3 months after delivery.

In the low-risk group one patient revealed a symptomatic deep vein thrombosis (0.4 %) and none with enoxaparin prophylaxis. There were no significant differences in minor or major bleeding complications between both groups.

The authors stated that their risk score model is safe and effective in avoiding LMWH prophylaxis in nearly 50 % of patients.

Recently, a Chinese group suggested a local risk score model for VTE prophylaxis after c.d. [33]. Major risk factors were overweight/obesity, maternal age ≥40 years, multiple pregnancy and postpartum hemorrhage.; 750 patients (87.3 %) were at low risk (VTE score ≤1) and were only advised for early mobilization after c.d., 109 (12.7 %) patients were at moderate risk (VTE score ≥2) had mechanical prophylaxis only and 28 (3.3 %) at high risk (VTE score ≥3) received mechanical and pharmacological prophylaxis for 10 days after delivery. The rate of emergency c.d. was 51 %.

None of the patients had postpartum VTE events. There was no information on the length of follow-up after c.d. The authors pointed out that applying the Royal College of Obstetricians and Gynaecologists guidelines to their cohort, 75.6 % patients would receive LMWH after c.d. which implies an overtreatment if the guidelines are followed.

An individual risk-stratified protocol on thromboprophylaxis after c.d. including 850 women was recently suggested by Kawaguchi et al. [34].

A total of 314 women without other risk factors received mechanical prophylaxis until discharge from hospital and 536 patients with one or more further risk factors additionally received LMWH prophylaxis: initially 5,000 IU UFH twice daily 6 h after c.d. followed by 20 mg enoxaparin twice daily for 5 days.

There were no symptomatic thromboembolic events during the follow-up of 4 weeks after delivery. Fewer bleeding complications were observed in patients with mechanical prophylaxis (2.2 %) compared to patients with LMWH prophylaxis (5.6 %; p=0.022). According to the authors opinion LMWH prophylaxis after c.d. stratified to VTE risk factors is effective and safe.

Recommendations from major guidelines

As highlighted by numerous authors during the last few years [35], [36], [37], [38], [39] guideline recommendations on post - cesarean thromboprophylaxis differ markedly in terms of which woman should receive pharmacologic thromboprophylaxis dependent on individual risk factors, which heparin should be given at which dosages and intervals and how long should prophylaxis be continued after c.d [40]. The risk threshold at which pharmacologic prophylaxis should be recommended is still a matter of debate. Guideline recommendations on pharmacologic thromboprophylaxis after c.cd. are mainly based on expert opinion and epidemiologic data rather than evidence-based results from RCT or other high-quality trials [41].

Appropriately powered RCTs to prove the efficacy of pharmacologic prophylaxis after c.d. are apparently not feasible [39]. Table 2 shows recommendations of major guidelines on thromboprophylaxis after c.d. A cohort study on 293 women attending a tertiary hospital in the United States has shown that 85 % of patients would need post -cesarean pharmacologic prophylaxis following the RCOG guidelines compared to 1 % of patients following the ACOG guidelines and 34.8 % of patients when applying the ACCP guidelines [35]. Federspiel et al. recently projected the implications of guideline adoption on the incidence VTE until the 6th after c.d. using administrative data from the 2015–2019 United States Nationwide Readmissions Database [15]. The American Society for Hematology guidelines [42] would avert 5 % of VTE cases, the ACCP guidelines 21 % and the RCOG guidelines 58 %, respectively. At the same time, LMWH usage would increase significantly from 87 to 115 doses per 1,000 c.d. considering the ASH guidelines to 7,233–38648 doses per 1,000 c.d., considering the RCOG guidelines. Besides the considerably higher costs with the more extensive use of LMWH the authors estimated a higher rate of wound complications when comparing ACOG guidelines (39.6/1,000 patients) to RCOG guidelines (58.4/1,000 patients).

Maternal mortality during the first 6 weeks after c.d. could potentially be reduced by universal LMWH prophylaxis from 238 (19.8/100,000) without prophylaxis to 217 cases (18.1/100,000) with LMWH prophylaxis necessitating 46 million doses of LMWH [15].

Discussion

Due to the lack of evidence optimal post-cesarean thromboprophylaxis remains an unsolved problem. Epidemiologic studies have shown that c.d. more than doubles the risk for VTE, which is even higher in women with emergency c.d. [10]. However, the increase in VTE and the benefit of any intervention to prevent VTE may be overestimated, when considering only results from large databases and retrospective analyses [36], 48]. Nevertheless, the risk of VTE following c.d. c.d. seems to be independent of other VTE risk factors [10]. The increased risk of VTE may be the rationale for the use of pharmacologic thromboprophylaxis in women with c.d. Decision-making for or against post-cesarean thromboprophylaxis is a common daily challenge for obstetricians and should be based on evidence-based recommendations. However, data from randomized controlled trials or other high-quality studies are very limited (Table.1). According to meta-analyses no significant differences were found in the risk of VTE, when comparing LMWH with no treatment, LMWH with UFH and LMWH among each other [25], 49].

The main problems in this context are the inadequate statistical power of studies due to small sample size and a considerable heterogeneity among studies regarding study design, risk profile of patients included type of c.d (elective/emergency), kinds and dosages of heparin used and duration of prophylaxis after delivery. Several risk assessment models to predict VTE after c.d. have been proposed, however they were not adequately validated by well-designed clinical studies and refined to support their `real life applicability [3], 50].

To prevent one VTE approximately 4,000 women must be treated, if LMWH is administered for one week following c.d. The basic risk of death from pulmonary embolism (PE) in the first week after c.d. has been estimated to be 0.4/100,000 women, postulating 70 % protection with LMWH, the number needed to treat to prevent one maternal death is approximately 360,000 pregnancies; assuming 70 % protection with LMWH, the number needed to treat to prevent one death is approximately 360,000 [9], 36]. The potential benefit of pharmacologic prophylaxis needs to be weighed against the potential for adverse outcomes associated with the intervention. LMWH prophylaxis has shown to increase the risk of bleeding or hematomas by 5–8.5 times compared with placebo, while the rates of blood transfusions and wound complications showed no significant differences in meta-analyses [25], 49]. It should be kept in mind, that a universal LMWH prophylaxis may lead to a tremendous increase in costs due to the high number of LMWH dosages needed, as shown by Federspiel et al. [15]. Friedman and D′ Alton recently discussed the pros and cons of an extended use of pharmacologic thromboprophylaxis in an expert review [39]. Due to lack of evidence, guidelines are mainly based on expert opinion being reflected in the low grades of recommendations (Table 2). Current guidelines do not recommend pharmacologic thromboprophylaxis in women with elective c.d., if no additional VTE risk factors are present. On the other hand, it may be a challenge for obstetricians to selectively exclude women from post-cesarean prophylaxis in view of the large number of risk factors and the problem to identify all these risk factors completely and timely (4,48). Hence, some authors advocated universal LMWH prophylaxis after c.d. rather than selective [4], 39], 51], 52]. In the case of elective c.d. and additional risk factors as well as in case of emergency c.d. most guidelines recommend pharmacologic thromboprophylaxis for 7–10 days after delivery, and at least 6 weeks in high-risk patients with persistent risk factors, particularly in women with inherited thrombophilia.

It has been shown that there is a considerable discrepancy between clinical practice in the use of VTE prophylaxis and guideline recommendations [17], [18], [19]. It can be assumed that the obstetrician`s clinical judgement may play a more important role than guideline recommendations in decision-making for or against post-cesarean thromboprophylaxis.

For example, in a German-wide survey 93 % of 389 respondents favored universal heparin prophylaxis after elective and emergency c.d, irrespective of additional VTE risk factors [19]. The Society for Maternal–Fetal Medicine [48] and the Consensus Bundle on Venous Thromboembolism from the National Partnership for Maternal Safety [52] recommended that each institution should develop a patient safety bundle with an institutional protocol for VTE prophylaxis among women who undergo c.d. (best practice). Randomized controlled trials with adequate statistical power are urgently needed to evaluate the efficacy of post-cesarean prophylaxis, however, it is debatable, whether such trials are feasible. Until there is no more evidence from such studies the decision for or against post-cesarean pharmacologic thromboprophylaxis may remain a case-by-case decision of the obstetrician based on his clinical judgement related to the patient individual VTE risk profile and the respective guideline recommendations followed, unless he prefers an universal pharmacologic thromboprophylaxis following c.d.