Approximately 32.2 % of adult men and 35.5 % of adult women are considered obese, and in 2009, almost 220,000 bariatric surgeries were performed in the United States [1, 2]. In the morbidly obese population that undergoes either open or laparoscopic bariatric surgery, venous thromboembolism (VTE) is the leading cause of morbidity and mortality. Deep vein thrombosis (DVT) and pulmonary embolism (PE) are major complications of bariatric surgery, with nearly 50 % of all deaths due to PE [3]. The increasing number of obese individuals coupled with the increased number of bariatric surgeries performed each year makes VTE an ever more important pathology.

The risk of VTE is faced by all patients who undergo bariatric surgery. However, several other factors place a patient at an even higher risk for the development of DVT/PE. The role of obesity as an independent risk factor for VTE is debated. Recent studies indicate that obesity places patients at a higher risk for VTE [4–7], and in addition to obese status, numerous other risk factors have been identified including a history of DVT/PE, a known history of a hypercoagulable disorder, severe immobility, estrogen medication, and advanced age [5, 8, 9]. This suggests that all bariatric surgery patients are at increased risk.

A study that surveyed members of the American Society for Bariatric Surgery in 2000 demonstrated that 86 % of the patients who had surgery performed by bariatric surgeons were considered to be at high risk for VTE [10]. However, the specific factors that place an obese patient in the high-risk category were not elucidated, and since then, no consensus has been formulated.

Due to the silent nature of VTE and the difficulty establishing a diagnosis of DVT/PE before the onset of symptoms, prophylaxis against VTE is paramount. A variety of treatment regimens for prophylaxis of VTE have been proposed. The majority of studies describe some form of pharmaceutic treatment coupled with the use of mechanical compression devices or mechanical compression devices alone [10–14]. However, no universally established protocol exists specifically for the high-risk bariatric patient.

We report the results of a survey encompassing members of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) who perform bariatric surgery. The respondents were asked to identify factors that put bariatric patients at high risk for VTE. The respondents also were asked to report the prophylactic treatment regimens they used to prevent DVT/PE in the high-risk bariatric patient group.

Materials and methods

Members of SAGES were contacted and asked to complete an online survey. The survey consisted of 26 questions regarding their current practices. The surgeons were asked to identify whether they performed bariatric surgery, their approach, and the total number of cases they managed annually. They were further queried on the preoperative risk factors they thought identified patients at high risk for VTE. Routine preoperative screening methods and prophylaxis patterns among high-risk patients were queried, including mechanical and pharmacologic prophylaxis. Members were asked to identify the chemoprophylactic agent together with initiation and duration of treatment. Their use of inferior vena cava (IVC) filters and postoperative screening also was queried. The results are described as percentages of completed surveys.

Results

Of the 385 surgeons who responded to the survey, 313 (81 %) reported themselves to be bariatric surgeons. The majority of the respondents were from the United States (75.8 %). They were stratified into three groups based on the number of bariatric cases they managed annually: fewer than 50 cases (25.6 %), 51–100 cases (31.6 %), or more than 100 cases (42.8 %). Most of those surveyed (98.7 %) routinely performed bariatric procedures laparoscopically. The respondents were asked about their practice patterns with regard to preoperative, operative, and postoperative VTE prophylaxis for high-risk patients.

Preoperative risk factors

The respondents identified the following individual factors as qualifying their patients to be considered at high risk for VTE: history of DVT/PE (99.7 %), known hypercoagulable status (99.7 %), severe immobility (98.6 %), body mass index (BMI) exceeding 55 kg/m2 (88.6 %), and partial pressure of arterial oxygen (PaO2) lower than 60 mmHg (65.9 %). A history of DVT/PE was identified by 69.9 % of respondents as the most significant single factor, followed by known hypercoagulable status (17.8 %), severe immobility (9.6 %), BMI exceeding 55 kg/m2 (1.4 %), and PaO2 lower than 60 mmHg (1.4 %) (Fig. 1). Nearly half of those surveyed (47.7 %) considered that three or more of these factors placed a patient at high risk for VTE compared with 7 % who considered high risk to be indicated by two factors. For 44.4 %, only one factor was necessary to identify a patient at high risk, whereas 1 % thought none of these factors placed a patient at high risk for VTE.

Fig. 1
figure 1

The most important single factor for determining VTE risk for bariatric patients

Full size image

More than half of the surgeons (56 %) routinely screened high-risk patients preoperatively for DVT: 33.1 % by clinical exam alone and 20.9 % with routine ultrasound. As reported, 46 % did not routinely screen high-risk bariatric patients for DVT. Only 23.2 % of the respondents routinely screened patients for previously unidentified hypercoagulable status.

Routine preoperative VTE chemoprophylaxis was applied by 92.4 % of the respondents (Fig. 2), with 48 % using unfractionated heparin, 33.4 % using Lovenox, 2.6 % using Fondaparinux, and 8.3 % using another anticoagulant. As reported, 28.1 % of the surgeons routinely used retrievable IVC filters preoperatively for high-risk bariatric patients, 0.3 % routinely used permanent filters, and 71.5 % did not routinely use IVC filters for high-risk patients.

Fig. 2
figure 2

Routine chemical prophylaxis for high-risk bariatric patients

Full size image

Operative and postoperative prophylaxis

Intraoperatively, 96.3 % of respondents placed sequential compression devices (SCD) on high-risk patients. As reported, 15 % used additional intraoperative chemical prophylaxis beyond the preoperative dose, with 7.3 % using additional heparin, 4.3 % using Lovenox, 0.7 % using Fondaparinux, and 2.7 % using other agents.

Postoperatively, 91.6 % of the surgeons used SCDs on their high-risk patients, with 97 % also using chemical prophylaxis. The most commonly used postoperative chemoprophylactic agent was Lovenox (49.5 %), followed by heparin (33 %), other agents (9.1 %), and Fondaparinux (5.4 %).

Postoperative chemoprophylaxis was started most commonly on postoperative day 0 (70 %), followed by postoperative day 1 (27.3 %). Chemical prophylaxis was discontinued at discharge in most cases (48.5 %). However, 43.8 % of the surgeons continued postoperative chemoprophylaxis for outpatients. The most common chemoprophylactic agent used for outpatients was Lovenox (39.7 %). The most common durations of outpatient therapy were 2 to 4 weeks (40.1 %) and fewer than 2 weeks (20.2 %).

Routine use of IVC filters for high-risk bariatric patients was reported by 28.1 % of the respondents, and 51.2 % of the surgeons routinely discharged patients using IVC filters coupled with chemoprophylaxis. The most common agent used was Lovenox (36.4 %), followed by other agents (8.8 %), heparin (3.7 %), and Fondaparinux (2.4 %). The most common duration of therapy was 2–4 weeks, followed by fewer than 2 weeks. The IVC filters were most commonly removed 30–90 days postoperatively (55.2 %). The remaining filters were removed after 90 days (23.2 %) or within 30 days (21.5 %).

Discussion

A significant percentage of morbidly obese bariatric patients are described as high risk by surgeons, but no consensus exists regarding which risk factors or what number of risk factors are required by physicians to identify a patient as high risk. Our study sought to identify these factors and also to describe current practice patterns regarding screening and the use of chemoprophylaxis and IVC filters.

Among the risk factors identified for considering patients at high risk (BMI > 55 kg/m2, history of DVT/PE, hypercoagulable status, severe immobility, and PaO2 < 60 mmHg), the majority of respondents considered a history of DVT/PE as the most important single risk factor. Whereas 44 % of surgeons needed only one of these risk factors to identify a patient at high risk, 47 % required a combination of three or more factors.

Previous VTE has been established as an independent risk factor for future VTE [15], and research has shown that among patients with previous VTE events, recurrent DVT/PE occurred for 33 % of patients who underwent bariatric surgery [16]. The same study demonstrated that among patients with no history of VTE, the incidence of subsequent DVT/PE was zero.

Our study found that 69.9 % of the respondents identified a history of DVT/PE as the most important risk factor for VTE, but only 54 % routinely screened high-risk patients preoperatively for DVT. As reported, 33.1 % of the surgeons used clinical exam alone, and 20.9 % used ultrasound. Overall, the majority of surgeons used a history of DVT/PE as an important marker to identify high-risk patients, but only 20 % screened patients with preoperative ultrasound.

The inconsistency between evaluation of risk and preoperative screening also was seen in patients with possible hypercoagulable status. Although 99.7 % of the surgeons considered a patient with known hypercoagulable status at high risk for VTE, only 23.2 % evaluated for a previously unrecognized hypercoagulable status.

Research has shown that up to 70 % of patients presenting with VTE have a demonstrable inherited or acquired thrombophilia. These include protein C and protein S deficiency, antithrombin deficiency, factor V-Leiden mutation, elevated factor VIII, antiphospholipid syndrome, and hyperhomocysteinemia [17]. Further research has shown that patients wishing to undergo bariatric surgery have high rates of both acquired and inherited thrombophilias [18].

Although no currently existing data links bariatric surgery patients with known thrombophilias to higher rates of DVT/PE, inference from previous research suggests that this could be the case. Surgeons should strongly consider preoperative assessment for the high-risk patient with inherited or acquired hypercoagulable status.

Severely obese patients, especially those with BMIs exceeding 55 kg/m2, tend to experience respiratory insufficiency, which consists of obesity hypoventilation syndrome (arterial PaO2 < 60 mmHg or arterial partial pressure of carbon dioxide [PaCO2] > 47), obstructive sleep apnea, or both [19].

Obesity hypoventilation syndrome was considered by 65.9 % of our respondents as a factor placing a bariatric patient in the high-risk category. Patients with respiratory insufficiency who undergo bariatric surgery have a mortality rate seven times higher than patients without respiratory insufficiency before surgery [20]. This further establishes low PaO2 as a significant preoperative risk factor.

The stratification of morbidly obese patients into a high-risk category should prompt a more aggressive VTE prophylaxis regimen. However, no prophylactic regimen is uniformly accepted, and no evidence-based approach has been defined. Although the overall incidence of VTE from bariatric surgery is 1–3 % and considered low, the presence of VTE among patients undergoing surgery likely is underreported [21]. Clinically evident PE may occur without prodromal signs or symptoms of DVT, so the opportunity for medical intervention may have already passed. Additionally, the increasing number of bariatric procedures performed each year likely will further increase the number of VTE events that occur.

Preoperative use of unfractionated heparin or its low-molecular-weight variant is well agreed upon as a mainstay of VTE chemoprophylaxis for all bariatric surgery patients [11, 12, 14, 15]. Most (92.4 %) of our respondents used preoperative chemoprophylaxis for high-risk patients. This chemoprophylaxis was continued postoperatively, with the majority (70 %) receiving the first dose on postoperative day 0. The duration of chemoprophylaxis was much more variable. Whereas 5 % of the surgeons discontinued chemoprophylaxis at ambulation, 48.5 % discontinued treatment at discharge. For high-risk patients, 43.8 % of the surgeons continued chemoprophylaxis on an outpatient basis for 2–4 weeks. Despite continuation of chemoprophylaxis for outpatients, only 47.4 % of the surgeons routinely screened high-risk patients postoperatively for DVT, with the majority of screening performed by physical examination alone.

Although previous studies have shown that high-risk patients are more likely to experience VTE formation for 3 months postoperatively and that continued chemoprophylaxis reduces the incidence of VTE [15, 22], no definitive practice pattern was identified by our study. Nearly half of the respondents continued chemoprophylaxis, whereas the other half did not, and the surgeons did not identify whether the patients discharged with medication were considered higher risk than the patients discharged without medication. Furthermore, no standardized postoperative screening method was used before discontinuation of chemoprophylaxis.

Even more aggressive treatment is used by some surgeons, with placement of IVC filters. The use of IVC filters has increased dramatically from 7 % of bariatric patients in 1997 to 55 % in 2007 [23]. Our survey demonstrated that 28.5 % of surgeons routinely use IVC filters in high-risk bariatric patients. Nearly all the filters placed (99 %) are the retrievable form (rIVCF). Although findings have shown the permanent form of IVC filter to be beneficial in preventing VTE for bariatric patients [24] in the short term, a long-term study has shown an increase in DVT frequency during a period extending to 8 years [25].

In the high-risk bariatric patient population, rIVCFs are shown to have a low incidence of DVT, no clinically evident PE, and the advantage of removal when the greatest risk to the patient subsides [26]. The literature is replete with data both supporting and refuting the use of IVC filters. However, our survey showed that the majority of surgeons do not use IVC filters routinely for high-risk bariatric patients.

Conclusions

Overall, our study describes practice patterns with respect to VTE identification, screening, and prophylaxis among high-risk bariatric surgery patients. Nearly all the surgeons (99.7 %) who responded to this survey agreed that high-risk patients are identified by a group of risk factors (history of DVT/PE, BMI > 55 kg/m2, hypercoagulable disorders, severe immobility, and PaO2 < 60 mmHg). However, only half (54 %) routinely screen for DVT preoperatively. Of these, the majority use clinical examination alone.

Preoperative chemoprophylaxis for high-risk patients was used by 92 % of the respondents, but the duration of chemoprophylaxis varied. Nearly half of them discontinued VTE chemoprophylaxis at discharge, with the remaining surgeons doing so at 2–4 weeks. As reported, IVC filters were not routinely used for the high-risk bariatric surgery population, and less than half of the responding surgeons routinely screened for DVT postoperatively.

Although certain practice patterns were apparent in this survey, the duration of treatment and screening varied significantly. Further studies are needed to determine whether this variability is clinically significant. Ideally, future prospective studies will compare screening and prophylaxis patterns in an effort to determine which regimen is the most efficacious and to establish an evidence-based prophylactic treatment for VTE as well as standardized screening for the high-risk bariatric patient.