A comparison of international clinical practice guidelines for postpartum venous thromboembolism prophylaxis

Abstract

Background

Venous thromboembolism is a leading cause of maternal death worldwide. The postpartum period is a time of particularly increased risk. International guidelines provide recommendations for when a woman should be offered thromboembolism prophylaxis, however they differ greatly in their criteria as to which women qualify for low molecular weight heparin (LMWH). The aim of this study was to determine the most common risk factors for women being recommended LMWH and compare the proportion of women who would qualify for postpartum LMWH according to four international guidelines.

Materials and methods

This cross-sectional study evaluated rates of postpartum LMWH prophylaxis by applying guideline recommendations from the American College of Chest Physicians (ACCP), the Royal College of Obstetricians and Gynaecologists (RCOG), the Society of Obstetricians and Gynaecologists of Canada (SOGC) and the Society of Obstetric Medicine of Australia and New Zealand (SOMANZ). Demographic, medical and obstetric risk factors for venous thromboembolism were identified for individual women from one regional and two tertiary maternity hospitals in New South Wales, Australia between February and October 2022.

Results

A total of 338 women were included in the analyses. By applying RCOG guidelines, 53.6% of women would have qualified for postpartum LMWH compared with 40.2% of women using SOMANZ guidelines, 37.3% using SOGC guidelines, and 8.3% using ACCP guidelines. The most common risk factors were caesarean birth, maternal age greater than 35 years, body mass index above 30 kg/m² and instrumental birth.

Conclusions

There are considerable differences in the rates of women receiving postpartum pharmacological venous thromboembolism prophylaxis when recommendations from different international guidelines are applied. These differences reflect the wide variation in guideline recommendations for the use of LMWH following birth.

Background

Venous thromboembolism remains one of the leading direct causes of maternal death in Australia [1] and the developed world [2, 3], accounting for 11% of all maternal deaths in Australia between 2011 and 2020. During pregnancy, the relative risk of venous thromboembolism increases four-to-five-fold [4, 5]. The incidence of venous thromboembolism has been shown to be raised throughout pregnancy, but the per-day risk is highest six-weeks postpartum [4, 6, 7], particularly during the first week [5, 8].

Pregnancy is a risk factor for venous thromboembolism due to various factors, including venous stasis, immobility, increased plasma levels of coagulation factors, and compression of the vena cava and pelvic vessels by the uterus [9]. Observational and population-based studies have shown high rates of risk factors among pregnant cohorts [8, 10]. Caesarean birth, increased maternal age, preterm delivery, multiple birth, and obesity are common risk factors for postpartum venous thromboembolism [8, 10, 11]. Prior venous thromboembolism has been demonstrated to be a common risk factor in the antenatal period [11].

Guidelines universally recommend the use of low molecular weight heparin (LMWH) in high-risk women over other forms of pharmacological thromboprophylaxis [12,13,14,15]. The criteria used to determine which women should be recommended pharmacological thromboprophylaxis by four international guidelines [the American College of Chest Physicians (ACCP), the Royal College of Obstetricians and Gynaecologists (RCOG), the Society of Obstetricians and Gynaecologists of Canada (SOGC) and the Society of Obstetric Medicine of Australia and New Zealand (SOMANZ)] are shown in Table 1. There is inconsistency in the indications and duration of LMWH use between the guidelines. For example, SOMANZ and RCOG recommend LMWH following emergency caesarean birth whereas SOCG and ACCP only recommend LMWH following caesarean section if other risk factors are also present. All guidelines recommend six weeks of LMWH in high-risk women, including those with a personal history of venous thromboembolism. Early mobilisation and avoiding dehydration are recommended for all women.

Table 1 Summary of international guideline recommendations for postpartum pharmacological thromboprophylaxis

A 2021 Cochrane review of 29 trials found there was a lack of evidence regarding the benefits and harms of thromboprophylaxis during pregnancy and in the postpartum period [16]. Hence, international guidelines rely on expert consensus and extrapolation of data from non-obstetric populations to determine which women qualify for pharmacological thromboprophylaxis during the antenatal period and puerperium. As such, international guidelines vary in their recommendations, both in terms of who should receive prophylaxis as well as the dosing and duration of therapy. Previous studies have demonstrated a wide variation in the proportion of women qualifying for thromboprophylaxis when different international guideline recommendations have been applied [17,18,19]. A single centre study by Palmerola et al. (2016) comparing three international guidelines for thromboprophylaxis following caesarean birth found between 1% and 85% of patients qualified for pharmacological thromboprophylaxis [17]. A study by Gassmann et al. (2021) in Geneva applied recommendations from the RCOG, ACOG and ACCP to vaginal and caesarean births and found considerable variation in the proportion of women classified as high risk, with 40.1%, 8.7% and 9.9% of all births qualifying for thromboprophylaxis, respectively [18]. Similarly, a study by O’Shaughnessy et al. (2019) located in a large single centre in Dublin of 21,019 women found 7% of women would qualify for LMWH thromboprophylaxis when ACOG recommendations were applied and 37% would qualify for LMWH using RCOG guidelines [19].

Previous studies have all been single centre studies at tertiary referral hospitals, which are likely to have a larger proportion of high-risk patients, with data for some risk factors used to determine which women would qualify for thromboprophylaxis also missing [18, 19]. Moreover, prior studies have not evaluated the rates of pharmacological thromboprophylaxis in an Australian obstetric population should widely accepted international guideline recommendations, including from an Australian-developed guideline, be applied [18, 19].

This study aimed to apply Australian and international guideline recommendations across three sites, including one regional hospital, to determine the most common risk factors for women being recommended LMWH and the proportion of women who would qualify for LMWH following vaginal and caesarean births, using internationally applicable guidelines.

Methods

A prospective cross-sectional study was performed at three hospitals in New South Wales, Australia between February and October 2022. The three hospitals included two tertiary maternity hospitals (Royal North Shore Hospital and Westmead Hospital) located in Sydney) and one regional hospital (Orange Base Hospital) located in rural New South Wales. Both tertiary maternity hospitals can provide care for women and babies with complex medical needs, whereas Orange Base Hospital being a regional hospital, is equipped to provide care to women who birth after 34 weeks of pregnancy (e.g., it does not have a neonatal intensive care unit). Royal North Shore Hospital, Westmead Hospital and Orange Base Hospital reported 2,660, 5,195 and 1,122 births, respectively in 2021. During the same period, New South Wales reported 99,316 births [20]. Women who had given birth at ≥ 20 weeks’ gestation were eligible to be included in the study. Eligible participants were identified by the nurse unit manager (NUM) of the postnatal ward. The NUM advised if any woman was unsuitable for face-to-face contact due to factors such as severe peripartum psychiatric illness or admission to the intensive care unit. Women who were unable to understand the English language consent form were ineligible for the study. Eligible participants were provided with a participant information statement and provided written consent prior to data collection.

A survey was developed by study investigators that included criteria for pharmacological thromboprophylaxis contained in the guidelines from SOMANZ, RCOG, ACCP, and SOGC [12,13,14,15]. This survey included demographic questions (e.g., maternal age, body mass index (BMI)), questions regarding previous births, medical and obstetric history (e.g., preeclampsia, gestational diabetes), family history of venous thromboembolism and thrombophilia, and specific venous thrombembolism risk factors included in the guidelines. Data were collected from a face-to-face interview and from the medical records on day two postpartum and entered into an online Research Electronic Data Capture (REDCap) [21] survey at each site.

Descriptive summary statistics were performed on the entire cohort. The algorithms described in four guidelines were then applied to the baseline data to determine which proportion of women would qualify for LMWH thromboprophylaxis. Analysis was performed using IBM SPSS Statistics for Windows, Version 25.0. IBM SPSS Statistics Version 26. Armonk, MY: IBM Corp and Microsoft Excel version 16.64.

Ethical approval for the study was obtained through the Northern Sydney Local Health District Human Ethics Committee (approval code: 2021/ETH12199).

Results

During the study period, 338 women were recruited to participate in the study. Of these, 232 (68.6%) gave birth in a tertiary centre, and 106 (31.4%) in a large regional centre. The mean age of women was 32 years (range 18–48 years, SD 5.0), with 78/338 (23.1%) women > 35 years. The frequency for each criterion used by the different guidelines to stratify women into a high-risk group, and thus needing pharmacological venous thromboembolism prophylaxis, is shown in Table 2.

Table 2 Frequency of criterion included in guidelines across all participants

The most common risk factors that qualified women for LMWH were caesarean birth, increased BMI and maternal age > 35 years. Caesarean birth occurred in 150/338 (44.4%) women. Of these 150 women, 79 (23.4%) occurred in an emergency setting where there was a risk to the mother or child, 71 (21.0%) were elective, and 46 (30.7%) of caesarean sections occurred during labour. Additionally, 30 (8.9%) of women had an instrumental delivery. Fifteen (4.4%) women were diagnosed with preeclampsia during their pregnancy and 45 (13.3%) gave birth prior to 37 weeks’ gestation. Thirty-seven (10.9%) women had premature rupture of membranes. The incidence of other high-risk factors was low.

A personal history of venous thromboembolism was present in four participants (1.4%), most commonly due to lower limb injury (two), lower limb injury and combined oral contraceptive pill (one) and smoking (one). There was a personal diagnosis of thrombophilia in three patients, one each of antiphospholipid syndrome, heterozygous factor V Leiden, and heterozygous prothrombin G20210A. Six (1.8%) women received long term anticoagulation and 31 (9.2%) women received anticoagulation at some point during pregnancy, this was often due to concurrent COVID-19 infection. Nineteen (5.6%) women had a family history of venous thromboembolism in a first-degree relative.

The number of women qualifying for pharmacological thromboprophylaxis is outlined in Table 3. By applying the recommendations of an Australian guideline endorsed by SOMANZ, 40.2% of women would qualify for postpartum LMWH. Using the RCOG and SOGC guidelines, 53.6% and 37.3% of women would be recommended LMWH following birth, respectively. When the ACCP guidelines were applied, 8.3% of women qualified for chemical thromboprophylaxis.

When considering only caesarean births, application of the guidelines found that a greater proportion of women would be recommended LMWH following birth (Table 4). Applying the RCOG guideline, 81.3% of women qualified for LMWH. Pharmacological thromboprophylaxis was recommended in 74% and 60.7% of caesarean sections applying the SOMANZ and SOGC guidelines, respectively. Application of the ACCP guideline resulted in 16% of women undergoing caesarean section qualifying for LMWH.

Table 3 Number and proportion of women qualifying for pharmacological thromboprophylaxis under various international guidelines

Table 4 Number and proportion of women who had caesarean section qualifying for pharmacological thromboprophylaxis under various international guidelines

Discussion

The results of this study demonstrate the different thresholds for postpartum LMWH thromboprophylaxis across four international clinical practice guidelines. Application of ACCP guideline recommendations resulted in the smallest proportion of women receiving LMWH thromboprophylaxis (8.3%), whereas application of RCOG guideline recommendations resulted in a more than six-fold greater number of women (53.6%) receiving LWMH following birth. A similar proportion of the obstetric population qualified for postpartum LMWH thromboprophylaxis when SOMANZ (40.2%) and SOGC (37.3%) guideline recommendations were applied to the same obstetric cohort. To our knowledge, this study is the first to compare the incidence of qualifying for venous thromboembolic prophylaxis using different international guideline criteria in vaginal and caesarean births, within an Australian cohort of women and including an Australian guideline.

Similar to previous studies [17,18,19], the most common risk factors for thromboembolism in an obstetric cohort were caesarean birth, maternal age greater than 35 years and BMI more than 30. All of these venous thromboembolic risk factors are increasing in prevalence [22, 23]. The prevalence of these major risk factors in the New South Wales population in general in 2021 are outlined in the annual NSW Mothers and Babies report. The incidence of maternal age > 35 was 27.4%, the incidence of caesarean section and instrumental delivery were 37.6% and 11.7% respectively. The prevalence of BMI greater than 30 was 16.1% in the general NSW population [20]. Moreover, these factors are additive, with women who are older and overweight more likely to experience other complications such as preeclampsia, preterm birth and birth by caesarean section, which themselves are risk factors for thromboembolic disease [23, 24]. Hence, the importance of considering the changing demographics and risk profile of the obstetric population when guidelines for thromboprophylaxis are designed and updated.

The differences in the proportion of women qualifying for LMWH thromboprophylaxis between the clinical guidelines can be largely explained by the baseline risk each guideline uses to determine which women should be offered LMWH. The ACCP guideline uses a baseline risk > 3% for a woman developing venous thromboembolism as the level at which a woman would be recommended pharmacological thromboprophylaxis, whereas the SOGC uses a venous thromboembolic risk > 1% to guide the recommended use of LMWH. Although the SOMANZ and the RCOG guidelines do not specify a baseline risk at which LMWH is recommended, the relatively high proportion of women qualifying for LMWH when applying the SOMANZ and RCOG recommendations in this study suggests that the baseline risk used in both these guidelines is closer to that of the SOGC guideline, rather than that of the ACCP guideline.

Clinical practice guidelines base their recommendations for postpartum thromboprophylaxis upon patient risk factors, however the evidence on which to base indications for LMWH use and duration of therapy after birth is limited. Recommendations are based not on incontrovertible evidence, but largely on expert opinion, observational studies, and extrapolation from non-obstetric populations. Indeed, an acknowledgement of this very low-grade evidence can be found in each of the guidelines. Although caesarean birth is a common indication for postpartum LMWH, a recent Cochrane review comparing heparin thromboprophylaxis with no treatment/placebo, based on four trials and only 840 women, found no difference in symptomatic thromboembolic events [16]. Similarly, the Cochrane review found heparin thromboprophylaxis following any type of birth (vaginal or caesarean) had no effect on thromboembolic disease when compared with no treatment. Women with a stillbirth in the current pregnancy have been identified as being at particularly high risk of having a venous thromboembolic event, with an approximately six-fold increased risk, which is even higher than that posed by caesarean birth during labour [25]. Surprisingly, only two of the four guidelines (RCOG and SOGC) include stillbirth as a risk factor for thromboembolic disease. There is also a lack of consensus among the four guidelines around the use of LMWH following emergency caesarean section. Both the SOMANZ and RCOG guidelines recommend LMWH for emergency caesarean alone, whereas the SOGC and ACCP guidelines require another major risk factor (e.g., medical comorbidity) in conjunction with the emergency caesarean to qualify for LMWH.

The use of LMWH as prophylaxis for venous thromboembolism is not without risk. Although the Cochrane review found that LMWH did not seem to increase the risk of postpartum haemorrhage or major bleeding following birth, the level of evidence was considered very uncertain, with only small trials and low numbers of events [16]. However, the same systematic review found there was a possible increase in non-major/minor bleeding with LMWH (relative risk (RR) 2.12, 95% confidence interval (CI) 1.15–3.93; 1 trail with 284 women). Furthermore, heparin thromboprophylaxis following caesarean birth was associated with increased bleeding complications (RR 5.03, 95% CI 2.49–10.18; 2 trials with 714 women). Current guidelines for clinical practice reflect the very uncertain evidence regarding the efficacy of LMWH in preventing thromboembolic disease and possible adverse effects in women at increased risk of venous thromboembolism. Rigorously conducted large-scale randomised controlled trials will provide highest quality evidence, however these trials are unlikely to be feasible given the low rates of thromboembolic disease and the extremely large numbers of women in each trial arm needed to show a difference in clinically important outcomes [16, 26]. In the absence of large-scale trial data, observational studies based on registry and population-linkage data [25] will help address the gaps in evidence around the efficacy and safety of LMWH following birth. Surveys of Australian, North American and German clinicians have shown that there is significant variation in clinical practice when guidelines are inconsistent [7, 27, 28], highlighting the need for evidence to improve guideline recommendations. Given our study has shown such wide variations in the rate of qualifying for thromboprophylaxis when using current guidelines, with their limitations as discussed above, an international body should be established to develop new consensus guidelines to advise best practice care.

The strength of this study is that, to our knowledge, it is the first study to include all risk factors for thromboembolism, with previous studies unable to collect data on all risk factors. A previous study by Palmerola et al. (2016) only considered women following caesarean section in their analysis [17], a study by Gassmann et al. (2021) did not include information on prolonged immobilisation [18], and a study by O’Shaughnessy et al. (2019) excluded some rarer risk factors which were not routinely reported [19]. Our study collected data across three centres, including a regional centre and two tertiary centres with high diversity of ethnic backgrounds. This has enabled comparison of guidelines in a cohort of women that is more reflective of the Australian population. Previous studies comparing guidelines [17,18,19] have been conducted in large tertiary centres which are likely to have a greater proportion of higher risk women. The inclusion of a regional centre may have led to the inclusion of a lower risk population not requiring tertiary-level care.

Our study adds to the literature which has previously shown that there is considerable variation in the proportion of women qualifying for postpartum thromboprophylaxis when different international guideline recommendations are applied. A key learning from the current study is that while it was possible to assess all the risk factors for thromboembolism described in the various guidelines, collection of the data using face-to-face interview and manual review of the medical records is extremely time consuming and labour intensive. Future studies would benefit enormously from the routine collection of thromboembolic risk factor data for all women during and after pregnancy being entered into an electronic medical record system. This would enable a large number of women to be studied in a timely manner and the prevalence of thromboembolic risk factors to be accurately measured. Being able to identify women who were administered LMWH and who developed venous thromboembolism using linked electronic datasets would also benefit future studies. Future research should focus on measuring the performance of the various international guidelines in preventing thromboembolic disease. Comparison of rates of venous thromboembolism in health systems that advocate the use of a particular guideline should be compared with other health systems that advocate an alternative guideline. These comparisons will provide invaluable information on the relative effectiveness of one guideline over another, particularly when our study has shown that the number of women receiving LMWH is more than six-fold greater when recommendations from the RCOG guideline are applied compared with the ACCP guideline.

A limitation of the current study was the use of face-to-face data collection. Study investigators visited maternity wards to collect data and hence not all women were able to be sampled, including women who were discharged on the same day as giving birth. These women are likely to be lower risk and have less complications, in part accounting for the higher proportion of women qualifying for LMWH under RCOG guidelines when compared with the studies by Gassmann et al. (2021) and O’Shaughnessy et al. (2019) [18, 19]. Unfortunately, data on venous thromboembolism risk factors for women who were discharged from hospital early (prior to day two postpartum) or excluded by the NUM were not available because the hospitals in this study did not undertake routine venous thromboembolism assessment and thrombosis risk factors were not routinely documented in the medical records. A further limitation of the current study is that the number of women who were discharged before the face-to-face interview or were excluded by the nurse unit manager was not collected. The limitations of the current study highlight the importance of future studies being designed to collect these missing data.

In summary, current recommendations for venous thromboembolism prophylaxis greatly differ across different international guidelines. There are considerable differences in the rates of women receiving postpartum pharmacological venous thromboembolism prophylaxis when recommendations from different international guidelines are applied. These differences reflect the wide variation in guideline recommendations for the use of LMWH following birth.