Review of BMI-Based Pharmacologic Protocols for the Prevention of Venous Thromboembolism in Bariatric Surgery Patients

Introduction

As the population of obese individuals in the United States has increased, so has the number of weight loss procedures. It is well known that bariatric surgery patients have an increased risk of venous thromboembolism (VTE) compared with the general population. ¹ VTE is a major cause of morbidity and mortality in this population. The incidence of VTE and postoperative hemorrhage in bariatric surgery patients in several large series ranges from 0.2% to 1.9% and 0.73% to 5.3%, respectively.^2–11

It is generally agreed among bariatric surgeons that patients should have some form of mechanical and pharmaceutical prophylaxis. However, there is no universally accepted standard of care for VTE prophylaxis in bariatric surgery patients. ¹² Accordingly, the main objective of this study was to examine the safety and effectiveness of three different enoxaparin dosage protocols in their ability to prevent VTE and minimize hemorrhage. Additionally, the evolution of these protocols will be reviewed.

Materials and Methods

A retrospective study of 364 bariatric surgery patients was performed at a 511-bed, level 1 trauma, community teaching hospital following review and approval by the Institutional Research Review Board. Due to the retrospective nature of the study, a waiver of patient authorization and consent was granted. Patients evaluated in the study, underwent one of four bariatric surgery procedures between 2002 and 2010. Patients were divided into three groups based on their enoxaparin dosing. Enoxaparin dosing categories were as follows:

Group 1 (n=192) 40 mg enoxaparin 1–2 h preoperatively. Postoperative dose 1 mg/body–mass index (BMI) (rounded to nearest 10 mg) every 12 h.

Group 2 (n=97) 1 mg/BMI enoxaparin (rounded to nearest 5 mg) 1–2 h preoperatively. Postoperative dose 1 mg/BMI (rounded to nearest 10 mg) every 12 h.

Group 3 (n=75) 1 mg/BMI enoxaparin (rounded to nearest 5 mg) 1–2 h preoperatively. Postoperative dose 1 mg/BMI (rounded to nearest 5 mg) every 12 h.

All postoperative doses of enoxaparin were continued for the duration of the hospital stay, regardless of group. Additional VTE prophylactic measures included sequential compression devices (SCDs) during and after surgery and early ambulation (within 6 h of surgery). Patients were discharged without pharmaceutical prophylaxis. Patients were excluded from the study if they did not fit into one of the dosing categories. Data for these patients were abstracted from paper and electronic charting systems. Demographic data included patient age, sex, weight, and BMI. VTE risk factor information included sleep apnea, presence of a known clotting disorder, known history of VTE, smoking, estrogen use, age greater than or equal to 40 years, reoperation, and length of hospital stay. The data were analyzed using descriptive statistics, ANOVA, two-sample t-tests, and chi-squared (χ²) tests, as appropriate, to examine differences between dosing regimens. SPSS version 14.0 (IBM, Armonk NY) was used for the analysis. Differences with a p-value of<0.05 were considered significant.

Results

A total of 364 patients were evaluated in the study. The population comprised 81% female patients and the average BMI of the entire study group was 48.4±8.18. The majority of patients presented with at least one risk factor for VTE in addition to their morbid obesity. Ten patients had a prior history of VTE. All procedures were performed by the same surgeon and included 314 laparoscopic Roux-en-Y gastric bypasses (LRGB), 2 open Roux-en-Y gastric bypasses, 16 laparoscopic sleeve gastrectomies, and 32 gastric banding procedures. Of all procedures, 99.5% were performed laparoscopically. Six patients received prophylactic inferior vena cava filter placement. Across the three dosage groups, the average length of stay was 3.5±1.35 days. Table 1 summarizes the characteristics of the study sample as well as the three dosing groups individually. The groups were similar in all noted parameters with the exception of age [F(2361=7.28, p=0.001] and procedure type [F(2361=4.74, p=0.01]. Using Tukey's HSD, age differences were identified between groups 1 and 2, and groups 1 and 3 with group 1 being the oldest sample of the three. Procedure type means were different between groups 1 and 3, and groups 2 and 3. Group 1 was the only group that included sleeve gastrectomies. Group 3 had more gastric bands and less gastric bypass procedures than group 1 or 2. Regarding VTE risk factors, differences in group means were found for patients who smoked (χ²=9.73, df=2, p=0.008) and those who presented with diabetes mellitus (χ²=7.57, df=2, p=0.023), hypertension (χ²=11, df=2, p=0.003), and age greater than or equal to 40 years (χ²=9.22, df=2, p=0.010). Post hoc analysis showed that smoking and hypertension means differed in groups 1 and 3, and groups 2 and 3, while age greater than or equal to 40 and diabetes mellitus means were different in groups 1 and 2, and groups 1 and 3.

Two patients (0.5%) in this study developed VTE complications after discharge (Table 2). One of these patients (group 3) underwent a laparoscopic gastric bypass, was readmitted for dehydration, and received a peripherally inserted central catheter (PICC) line for hydration and nutrition. The patient was later found to have an upper extremity DVT associated with the PICC line 46 days after the original procedure. The other patient (group 1) underwent open conversion of a prior vertical banded gastroplasty to a gastric bypass in conjunction with a repair of a large ventral hernia with a biologic tissue mesh. The patient developed a wound infection requiring extensive debridement and a PE 31 days after the original procedure. Both patients were superobese (BMI >50) and had multiple comorbidities, including hypertension and sleep apnea. Both patients responded well to anticoagulation therapy and were released without further incident.

Overall, postoperative hemorrhage was suspected in 36 (9.8%) patients with 8 (2.2%) patients requiring transfusion postoperatively. No patients required intraoperative transfusion. Reoperation for hemorrhage was required in 5 (1.3%) patients. VTE occurred in 2 (0.5%) patients. There was not a significant difference in the incidence of transfusion, reoperation for hemorrhage, or VTE between groups 1, 2, or 3. Mortality (30-day) was 0% for all groups.

Of interest, there were two anastomotic leaks (0.5%; p=0.02). Both were in dosing group 3. One patient was a complex laparoscopic conversion of a Nissen fundoplication to a gastric bypass and was treated with percutaneous drains. The second was a superobese male laparoscopic gastric bypass patient and the leak was treated operatively. Both leaks originated from the gastrojejunostomy, and both patients recovered uneventfully.

Discussion

VTE is a leading cause of mortality among bariatric surgery patients. There is no single universally accepted pharmacologic prophylaxis regimen for this patient population. Published regimens for pharmacologic thromboembolism prophylaxis vary from no prophylaxis^3,7 to standard general surgery prophylaxis or fixed dosing,^4,6,13 to BMI-based regimens,^2,11 to continuation of prophylaxis after discharge,^8,9 as well as unconventional therapies such as dextran. ⁵ Table 3 provides a summary of previously reported prophylaxis regimens using enoxaparin and their associated VTE outcomes if reported. The objective of the study was to compare the safety and effectiveness of three VTE prophylaxis regimens.

Our VTE prophylaxis protocol over time encompassed three different dosing regimens of enoxaparin. The evolution of our protocol requires explanation. The initial protocol was based on the prevailing idea that standard general surgery prophylaxis was not adequate for bariatric surgery patients and was in line with the group 3 dosing regimen. This protocol was changed when the pharmacy staff commented on the difficulty and practicality of rounding enoxaparin to the nearest 5 mg on the bariatric ward. Thus, the prophylaxis was simplified to the dosing applied to group 2. With increased experience with this regimen, there was subjectively more intraoperative oozing from staple lines requiring the additional time and effort of oversewing staple lines. In addition, there were three patients who required reoperation for hemorrhage in the perioperative period in group 2 and these events occurred close temporally. This resulted in a change to group 1 dosing that resulted in a decreased preoperative enoxaparin dose for most of the patients. A subjective decrease in intraoperative staple line bleeding was realized with this modification.

The current enoxaparin dosing regimen (group 1) appears to balance the concern for postoperative VTE with the intraoperative issue of bleeding staple lines and postoperative hemorrhage. Both the patients who developed VTE in this study were diagnosed after their initial hospitalization, after the development of a complication, and well after the cessation of their prophylaxis. With utilization of our current dosing regimen (group 1), the incidence of VTE (1/192 or 0.52%) and reoperation for hemorrhage (2/192 or 1.0%) are comparable to those published in the literature. The incidence of VTE and postoperative hemorrhage in bariatric surgery patients in several large series ranges from 0.2% to 1.9% and 0.73% to 5.3%, respectively.^2–11

Winegar et al. ¹⁰ reported the incidence of VTE events within 90 days of surgery in 73,921 bariatric surgery patients by analyzing the Bariatric Outcomes Longitudinal Database. This is a prospective database of patients undergoing bariatric surgery at centers participating in the Bariatric Surgery Center of Excellence program. The data are self-reported by surgeons and/or their representatives. Of all patients, 93.4% received some type of VTE prophylaxis whether pharmacologic or mechanical. The overall risk of VTE for all bariatric procedures within 90 days after surgery was 0.42%. The VTE risk for gastric bypass (0.55%) and sleeve gastrectomy (0.63%) is similar to that in our current dosing group (0.52%). However, 7% of procedures in our current dosing group were gastric bands, which have a lower risk for VTE (0.16% according to Winegar).

In 2012, Becattini ¹³ reported a meta-analysis that evaluated 3991 laparoscopic bariatric surgery patients in 19 studies. In 12 studies, the weighted mean incidence (WMI) of pulmonary embolism in patients undergoing laparoscopic bariatric surgery was 0.5% with unfractionated heparin or low molecular weight heparin. The WMI of bleeding was 3.6%. In three high-quality studies, the WMI of screened VTE with different regimens of heparin prophylaxis was 2.0%. In four studies, the WMI of symptomatic VTE was 0.6% and that of major bleeding was 2.0% with weight-adjusted doses of heparin prophylaxis.

Bariatric surgery patients range from those with a BMI of 35 to those with a BMI of 80 and above. This has led to the idea that a single fixed dosing regimen of pharmacologic VTE prophylaxis is not optimal for every patient in that wide BMI range. This is the rationale for BMI-based dosing of VTE pharmacologic prophylaxis. Miller and Rovito ² reviewed 255 LRGB patients who received BMI-based dosing of subcutaneous unfractionated heparin (SQH). Five thousand international unit subcutaneous (SC) were given for a BMI <50 and 7500 IU for a BMI >50 preoperatively. SQH was continued postoperatively at the same dose every 8 h until discharge. 1.2% developed VTE. There were 6 (2.4%) postoperative bleeding episodes, but no patient required reoperation for hemorrhage. Singh et al. ¹¹ reviewed 170 LRGB patients who received BMI-based enoxaparin. Patients were given a single dose of enoxaparin within 1 h before incision followed by twice daily administration postoperatively. Patients with BMI <40 received 30 mg enoxaparin, BMI 41–49 received 40 mg, BMI 50–59 received 50 mg, and BMI >59 received 60 mg within 1 h of surgery. The protocol was modified to exclude the 50 mg as prefilled syringes are dispensed in 30, 40, and 60 mg doses. All patients utilized SCDs and ambulation on postoperative day 1. No patient developed VTE. One (0.58%) patient required reoperation for hemorrhage.

Other studies have evaluated fixed VTE prophylactic doses. Some have included continued prophylaxis after discharge. The results are comparable to BMI-based dosing. Brasileiro et al. ⁶ prospectively evaluated 136 gastric bypass patients (69 laparoscopic and 57 open). Patients with a prior history of VTE were excluded. Enoxaparin 40 mg SC was administered after the beginning of the procedure and masintained for 15 days after surgery. SCDs were not used. Doppler ultrasound of the lower extremities was performed preoperatively and on the second and fifth week after surgery. One hundred twenty-six patients concluded full evaluation. Ten patients did not conclude full evaluation for the following reasons: 6 patients cancelled surgery, 1 had a heparin allergy, 1 had prior DVT, 1 was lost to follow up, and 1 died 7 days postoperatively due to bypass stomach hemorrhage causing rupture, sepsis, and death. One patient (0.79%) developed a DVT. Four cases (3.2%) of postoperative hemorrhage were noted, but treatment (reoperation, transfusion) was not reported. Mortality was 0.79%. Escalante-Tattersfield et al. ⁴ reviewed 618 LRGB patients who received 5000 IU SQH preoperatively and postoperatively every 8 h for 24 h. Patients were then started on enoxaparin 40 mg every 12 h until discharge. 0.2% developed DVT and no patient required reoperation for hemorrhage. Raftopoulos et al. ⁸ reviewed 308 patients divided into two pharmacologic prophylaxis groups who underwent various bariatric procedures. Group A received enoxaparin 30 mg SC 1 h preoperatively followed by 30 mg twice a day until discharge. Group B received enoxaparin 30 mg SC twice a day after surgery until discharge and then received 40 mg daily for 10 days after discharge. All patients utilized SCDs. All patients underwent bilateral lower extremity doppler ultrasound evaluation on the day of hospital discharge. There were three DVT and three PE for a total VTE incidence of 1.9%. All occurred in group A, which was significant. Reoperation for hemorrhage was required in two patients (0.6%), one from each group. Mortality was 0%.

It is debated by some whether pharmacologic VTE prophylaxis is necessary at all. Some studies report on no pharmacologic VTE prophylaxis also with comparable results to BMI-based and conventional general surgery VTE prophylaxis regimens. Gonzalez et al. ³ reviewed 380 LRGB patients who received no pharmacologic VTE prophylaxis. SCDs were used intraoperatively and postoperatively until routine early ambulation. 0.26% developed DVT. Reoperation for hemorrhage and mortality was not evaluated. The conclusion was that pharmacologic prophylaxis is not mandatory when calf-length SCDs are used and with a relatively short operative time. Mean operative time was 103.3 min. Clements et al. ⁷ reviewed 957 LRGB patients who had no prior history of VTE. No pharmacologic prophylaxis was used. All patients utilized SCDs and ambulated on the day of surgery. Three patients (0.31%) developed clinically evident DVT and one patient (0.10%) had a PE. Two patients (0.2%) required reoperation for hemorrhage. Mortality was 0.10%.

There are also unconventional therapies for VTE prophylaxis that have been shown to be effective. Westling et al. ⁵ prospectively evaluated 116 gastric bypass patients ultrasonographically. Preoperative ultrasound was performed as well as postoperative examinations on postoperative days 3 and 6. Eighty-six procedures were performed through laparotomy, 30 were performed laparoscopically. A single dose of dextran was given for VTE prophylaxis preoperatively in the majority of the patients. VTE occurred in three patients (2.5%). One (0.9%) was a symptomatic PE and 2 (1.7%) were asymptomatic DVT. One additional patient had a proximal saphenous vein thrombosis, which is technically not a DVT. Reoperation occurred in 1.7%. Indication for reoperation was not reported.

There are many acceptable dosing regimens that have been reported. The optimal VTE dosing regimen has yet to be determined.

This study has several limitations. Data were collected and analyzed retrospectively and the study was not randomized. The sample sizes may be too small to detect important clinical differences in dosing regimens. Twenty-nine patients from the bariatric surgery panel were excluded as they did not fit into any of the defined dosing regimens. This may be because they were at higher risk for VTE than the average patient and may have been treated differently. Excluding this number of patients relative to the sample size may have introduced sampling bias into this study.

As noted previously, there were two anastomotic leaks in group 3, which is statistically significant. Group 3 was the initial group in the laparoscopic bariatric experience of our program. The increased incidence of leaks could be accounted for based on this fact and the effect of the learning curve and not necessarily on enoxaparin dosing.

Conclusion

Comparison of three VTE prophylaxis dosing regimens revealed no differences in safety and effectiveness. Further, the incidence of bleeding complications and VTE noted in this study is in line with previously reported studies. The current VTE prophylaxis regimen of enoxaparin 40 mg SC 1–2 h preoperatively followed by 1 mg/BMI (rounded to the nearest 10 mg) SC every 12 h until discharge coupled with mechanical VTE prophylaxis and early ambulation appears to be safe and effective in bariatric surgery patients.