Portaldacirurgiaplastica.com.br

Incidence of Hematoma Complication withHeparin Venous Thrombosis Prophylaxis afterTRAM Flap Breast Reconstruction Background: Randomized controlled studies provide ample evidence that hep-
arin is effective in reducing the risk of thromboembolic complications. Never- theless, plastic surgeons are often reluctant to use heparin chemoprophylaxis for fear of postoperative bleeding. The authors investigated whether heparin chemoprophylaxis was associated with postoperative hematoma that required evacuation in patients who underwent transverse rectus abdominis myocutane- ous (TRAM) flap breast reconstruction.
Methods: A multicenter retrospective review of consecutive TRAM flap cases
identified 679 patients, 392 in the heparin-treated group and 287 in the control
group. The post hoc sample sizes were adequate to detect a 5 percent difference
in hematoma rate with 89 percent power at an alpha level of 5 percent (p Ͻ 0.05).
Outcome measures of reoperative hematoma, deep vein thrombosis, and pul-
monary embolism were recorded.
Results: Reoperative hematoma occurred in 0.5 percent of patients in the
heparin-treated group and 1.0 percent of patients in the control group; this
difference was not statistically significant (p ϭ 0.66). Thromboembolic events
were detected at a low rate (0.8 percent in the heparin-treated group versus 1.4
percent in the untreated group; p ϭ 0.46).
Conclusions: The use of heparin for venous thrombotic prophylaxis did not
increase the risk of reoperative hematoma after breast reconstruction with
abdominal tissue. The authors propose a risk assessment that balances a statis-
tical hematoma rate of 0.5 to 5 percent (clinically observed rate, 0.5 percent)
with use of heparin prophylaxis against a rare (clinically observed rate, 1.4
percent) but morbid occurrence of thromboembolic complications when che-
moprophylaxis is omitted.
(Plast. Reconstr. Surg. 121: 1101, 2008.) Deepveinthrombosisisdetectedatasignif- aggregategeneralsurgerypopulation.5,6However,
icant rate of 10 to 40 percent in patients given the significant morbidity and possible mor- after general surgical operations, when ex- tality of venous thrombotic complications, the pre- amined in screening studies.1–4 Pulmonary embo- vention of such adverse events has been the focus lism contributes to 10 percent of all hospital of patient safety initiatives worldwide.
deaths and 5 percent of perioperative mortality.
The incidence of clinically evident venous throm- cations after plastic surgery have been reported botic event is low, occurring in 4 to 7 percent of mostly from survey studies and small case series.
patients after hip fracture fixation, 3.5 percent Surveys of plastic surgeons reported that clini- after bariatric surgery, and 0.1 to 0.8 percent of cally significant thromboembolic complicationoccurred in 0.49 percent of patients after rhyt- From Harvard Medical School, Brigham and Women’s Hos- idectomy, 1.7 percent after large-volume (5 li- pital, Massachusetts General Hospital. ter) suction lipectomy, and 0.6 to 2 percent after Received for publication April 23, 2007; accepted June 5, abdominoplasty.7–10 Reports of thrombotic com- 2007.
Presented at the 86th Annual Meeting of the AmericanAssociation of Plastic Surgeons, in Coeur d’Alene, Idaho,May 22, 2007. Disclosure: The authors have no commercial asso-
Copyright 2008 by the American Society of Plastic Surgeons ciations or financial interests to disclose. www.PRSJournal.com
Plastic and Reconstructive Surgery • April 2008 plications in clinical series ranges from 1.1 to 2.3 a retrospective review of consecutive cases of percent after abdominoplasty to 9.4 percent af- TRAM flap breast reconstruction was performed.
ter belt lipectomy.11–13 Obesity and the combi- The TRAM flap procedure is the most commonly nation of abdominoplasty with intraabdominal performed operation for autologous breast recon- procedures have been cited as risk factors for struction at our institutions and would yield the development of venous thrombosis after exci- largest patient pool. Cases of other types of autol- ogous breast reconstruction such as latissimus dorsi flap or perforator flaps were excluded. Pedi- studies provide ample evidence that the use of cled TRAM and free TRAM flap procedures were unfractionated or low-molecular-weight hepa- all performed essentially as described previously, rin is effective in reducing the risk of thrombo- with minor technical variability among surgeons, embolic complications.14–22 The most frequently such as method of fascial closure, numbers and referenced guideline on prevention of venous types of drains used, and suture types.26–28 In cases of free TRAM flap procedures, microvascular ar- College of Chest Physicians, which stratifies risk terial anastomosis was hand sewn with 9-0 nylon of venous thrombosis and recommends modal- sutures and venous anastomosis was performed ities of prophylaxis.20 Several reviews in plastic with a vessel coupling system (Synovis, Inc., Bir- surgery have examined these guidelines and of- mingham, Ala.), where the recipient vessels were either the internal mammary or the thoracodorsal There is a paucity of reports in the plastic artery and associated venae comitantes.
surgery literature on the incidence of hematoma Cases were collected from January of 2000 to complication with the use of heparin for venous June of 2006, encompassing immediate or delayed thrombosis prophylaxis. The only study in the En- breast reconstruction, with 679 cases from 14 sur- glish literature is a recent retrospective review of geons performed at Brigham and Women’s Hos- 126 rhytidectomy cases from a single surgeon, pital and Massachusetts General Hospital (Table where a 16.2 percent rate of postoperative bleed- 1). None of the patients were treated with subcu- ing was reported when low-molecular-weight hep- taneous heparin before surgery. All of the patients arin was used perioperatively for thromboembolic received mechanical thromboembolic prophy- prophylaxis.24 The need for education of practic- laxis in the form of elastic compressive stockings ing plastic surgeons on the risks and benefits of and sequential pneumatic compression boots, deep vein thrombosis prophylaxis is made evident which were placed before induction of general in a recent survey study that found consistent use anesthesia. The compressive stockings and pneu- of any form of deep vein thrombosis prophylaxis matic boots were worn by patients until they were in only 43.7 percent of liposuction and 60.8 per- ambulatory in the postoperative period. Whether cent of combination lipectomy procedures.25 the patient received heparin for venous thrombo- sis prophylaxis correlated with the practice of the review of 679 consecutive patients who under- resident staff that cared for the patient postoper- went transverse rectus abdominis myocutaneous atively. In free TRAM flap cases (n ϭ 61), all pa- (TRAM) flap breast reconstruction to study whether tients also received an intravenous bolus of hep- the use of heparin for thromboembolic prophy- arin (2500 units) before microanastomosis and laxis increased the risk of postoperative bleeding.
placed on aspirin (300 mg per rectum immedi- As a secondary outcome, we examined whether ately after surgery and then 325 mg daily there- a difference in the rate of clinically significant after for 1 month). Patients who underwent bilat- thromboembolic complication can be detected eral free TRAM flap surgery (n ϭ 10) received an between the treatment and control groups. Wechose the breast reconstruction patient popula-tion as the model for our study, as these patients Table 1. Multicenter Consecutive Series of TRAM
have thorough preoperative evaluation and are Flap Breast Reconstruction
followed long term. This study provides evidence Flap Type
Heparin-Treated
to support plastic surgeons in their risk-to-benefit analysis of pharmacologic prophylaxis.
PATIENTS AND METHODS
After obtaining approval from the clinical re- search internal review boards at each institution, Volume 121, Number 4 • Hematoma Complication with Heparin intravenous bolus of heparin (2500 units) for each heparin (three cases), or developed postoperative flap, for a cumulative 5000-unit dose. When che- hematoma before heparin was administered (one moprophylaxis was used, unfractionated heparin case). Cases notable for reoperative hematoma was used in almost all cases, with 5000 units in- were examined in detail, noting the total operative jected subcutaneously twice daily. Included in the time, location of hematoma (chest recipient site treatment group are 17 patients who received the or abdominal donor site), timing of reconstruc- low-molecular-weight heparin dalteparin (Frag- tion (immediate or delayed), timing of hematoma min; Pfizer, Inc., New York, N.Y.), 2500 units twice (postoperative day), transfusion requirement, and daily. Subcutaneous heparin was administered un- whether the hematoma led to compromise of flap til the patient was discharged from the hospital.
viability. Cases remarkable for development of No patient received heparin prophylaxis after dis- deep vein thrombosis or pulmonary embolism were examined further, noting the patient’s age, Records of each case were reviewed in detail, body mass index, smoking history, operative time, with particular attention to preoperative history and physical examination, oncology clinic evalu- The t test and Pearson’s chi-square analysis ation, pathology report, operative notes, dis- were used to compare patient characteristics be- charge summaries, and imaging studies. Patient tween the treatment and control groups. Statistics characteristics of age, body mass index, operative of group comparisons are listed as means with SD time, smoking status, comorbid conditions (hy- in parentheses (Table 2). The Fisher’s exact test pertension, diabetes, coronary disease), and peri- was applied to determine statistical significance of operative cancer stage were noted. Adverse com- outcome measures between the treatment and the plications of hematoma formation were determined control groups. Our post hoc power analysis based by noting which patients returned to the operating on these sample sizes allowed us to detect a 5 room for hematoma evacuation (reoperative hema- percent difference in hematoma rate with 89 per- toma). As all patient encounters are logged in the cent power at an alpha level of 5 percent (p Ͻ computerized longitudinal medical record of the 0.05). All statistical analyses were performed using Massachusetts General Hospital and Brigham and Stata, version 9.1 (Stata Corp., College Station, Women’s Hospital (Partner’s Healthcare System), we reviewed all available discharge records, oper-ative reports, and imaging studies to detect anybleeding or thrombotic complication that may have occurred after the initial TRAM flap proce- A multicenter retrospective review of consec- dure. However, our chart review would not detect utive cases in which abdominal tissue was used for cases of small hematomas that did not result in re- breast reconstruction was carried out (Table 1).
operation or notation in the discharge summary.
We identified 679 patients who underwent imme- Presence of deep vein thrombosis or pulmo- diate or delayed breast reconstruction with pedi- nary embolism was determined by review of dis- cled TRAM (n ϭ 618) and free TRAM (n ϭ 61) charge summaries and review of whether radiol-ogy imaging was obtained in workup for deep vein Table 2. Patient Characteristics
thrombosis (lower extremity ultrasound) or pul-monary embolism (helical chest computed to- Heparin-
mography with angiographic reconstruction) dur- p
ing or after discharge from the TRAM flap procedure hospitalization. Therefore, thrombotic complications were detected when the patient ex- hibited signs or symptoms suspicious for deep vein thrombosis or pulmonary embolism that war- ranted further workup. Asymptomatic postopera- tive patients were not screened for deep vein A total of 685 TRAM flap cases were captured in the time period of the study, with six cases excluded, as the patients had a known bleeding or thrombotic disorders (two cases), were anticoag- ulated perioperatively on warfarin or intravenous Plastic and Reconstructive Surgery • April 2008 Table 3. Postoperative Heparin Pharmacologic Prophylaxis for Venous Thrombosis Does Not Increase
Hematoma Complication Rate*
Chemoprophylaxis
No Hematoma
Hematoma
Hematoma Rate (%)
Table 4. Postoperative Hematoma Complication Subgroup Characteristics
Location
Reconstruction
Compromise
BMI, body mass index; OR, operating room; POD, postoperative day; PRBC, packed red blood cells.
Table 5. Incidence of Detected Thromboembolic
The abdominal donor site was also explored in Complication after TRAM Flap Breast
patient 4, but no hematoma was found. In all cases, Reconstruction*
hematoma evacuation took place during the ini- Thrombotic
tial hospital course. Patient 5 required a one-unit Chemoprophylaxis DVT/PE DVT/PE Total
blood transfusion because of lightheadedness.
None of the patients became hemodynamically compromised or required escalation to a higher DVT, deep vein thrombosis; PE, pulmonary embolism.
low rate in our case series, 0.8 percent in theheparin-treated group versus 1.4 percent in the flaps. Of these patients, 392 received deep vein untreated group (Table 5). Clinical signs and thrombosis chemoprophylaxis with subcutaneous symptoms that led to workup studies detecting heparin postoperatively, whereas 287 patients did pulmonary embolism included dyspnea, tachycar- not receive chemoprophylaxis. Patient character- dia, and increased oxygen requirement. Clinical istics were well matched between the treatment signs and symptoms that led to workup studies and control groups, without any statistically sig- detecting deep vein thrombosis included asym- metric leg swelling and leg pain. None of the pa- We analyzed the case series for outcome mea- tients became hemodynamically unstable or re- sures of hematoma requiring operative evacuation quired transfer to a higher acuity level of care.
and development of thromboembolic events. We Therapeutic heparinization was achieved (either found that hematomas requiring operative explo- intravenous heparin for goal prothrombin time, ration occurred in two patients (0.5 percent) in or low-molecular-weight heparin dosed by weight) the heparin-treated group and three patients (1.0 as a bridge to anticoagulation with warfarin long percent) in the control group (Table 3). The oc- term (3 to 6 months). Fisher’s exact test did not currence of reoperative hematoma was not statis- demonstrate statistical significance for the differ- tically significant when analyzed with Fisher’s ex- ence of thromboembolic rate between the hepa- rin-treated and untreated groups (p ϭ 0.46).
In all five cases where a hematoma required The mean age of patients with clinically sig- operative exploration, the hematoma was found nificant thromboembolic event was 55 years, with in the chest TRAM flap recipient site (Table 4). In mean body mass index of 25, and the mean op- patient 1, distinct bleeding vessels were noted on erative time was 339 minutes; none of these pa- the chest wall and the TRAM flap, but no discrete rameters was statistically different from the base- bleeding sites were identified at the time of he- line patient profiles (Table 6). Six of the seven matoma evacuation in the remainder of patients.
patients (86 percent) with thromboembolic com- Volume 121, Number 4 • Hematoma Complication with Heparin Table 6. Thromboembolic Complication Subgroup Characteristics
OR Time (min)
Flap Type
DVT, deep vein thrombosis; PE, pulmonary embolism; BMI, body mass index; OR, operating room.
plications came from the pedicled TRAM flap when low-molecular-weight heparin was used in group. None of the patients treated with heparin the facial rhytidectomy patient population.24 It is developed heparin-induced thrombocytopenia important to point out that in the Durnig and Jungwirth study, low-molecular-weight heparinwas used, where the first dose was administered 2 DISCUSSION
hours before surgery. In our series, unfractionatedheparin was administered postoperatively (except We present the first large case series of major in 17 patients, where low-molecular-weight hepa- reconstruction in plastic surgery that investigates rin was used). Furthermore, development of he- the incidence of bleeding complications associ- matoma in a facial rhytidectomy patient may be ated with heparin chemoprophylaxis. Our retro- more apparent than a hematoma that forms under spective study suggests that the use of heparin for a mastectomy flap, TRAM flap, or TRAM abdominal venous thrombotic prophylaxis did not increase donor site. Therefore, our results may be more ap- the risk of reoperative hematoma after TRAM flap plicable to plastic surgical procedures of the trunk breast reconstruction. We propose a risk assess-ment that balances a statistical hematoma rate of 1 to 5 percent (clinically observed rate, 0.5 per- Our study failed to demonstrate a statistically cent) with use of heparin prophylaxis against a significant difference in the rate of thromboem- rare (clinically observed rate, 1.4 percent) but bolic complication between the heparin-treated morbid occurrence of thromboembolic compli- versus untreated groups. Given the expected low cations when chemoprophylaxis is omitted.
incidence of clinically significant thromboem- The frequency of bleeding complications af- bolic events (0.5 to 2 percent based on the plastic ter heparin chemoprophylaxis has been studied surgery literature), a sample size of 3300 would be in other surgical specialties.29–31 A recent meta- needed in each of the treatment and control analysis of 33 randomized controlled trial stud- groups to provide 90 percent power to detect a 1 ies using unfractionated heparin and low-molec- percent difference in thromboembolic rate. The ular-weight heparin in general surgery found a sample sizes in our study do not provide adequate wound hematoma rate of 5.7 percent and conse- quent reoperation rate of 1 percent.32 All random- Complications following pedicled and free ized controlled trial studies initiated heparin che- TRAM flap breast reconstructions have been re- moprophylaxis preoperatively, with mechanical ported, including hematoma and thromboem- prophylaxis initiated intraoperatively and heparin bolic events. In a series of 718 patients who un- continued postoperatively. Several meta-analyses derwent free TRAM flap breast reconstruction, found low-molecular-weight heparin to be associ- Chang and colleagues report rates of hematoma ated with a slightly higher risk of bleeding com- in the TRAM flap and abdominal donor sites of 1 plication than unfractionated heparin.30,31 Our re- percent and 3.2 percent, respectively.33 Schuster- sults agree with the randomized controlled trial man et al. found a 2 percent hematoma rate and studies in reporting a low rate of wound hema- no deep vein thrombosis in 163 patients who un- derwent free TRAM breast reconstruction.28 In a The only other study in the plastic surgery free TRAM flap series reported by Wang et al., the literature that examined the incidence of postop- deep vein thrombosis rate was 10.5 percent in erative bleeding complication with heparin che- patients with a body mass index over 30 (two of 19 moprophylaxis reported a statistically significant patients) and 1.1 percent in patients with a body 16.2 percent rate of postoperative hematoma mass index less than 30 (one of 88 patients).34 Plastic and Reconstructive Surgery • April 2008 Unfortunately, these reports did not specifically cancer patients) or obesity (20 percent over ideal address thromboembolic prophylaxis, such that body weight) elevate the patient into the highest methods of prophylaxis (i.e., mechanical, chemi- risk category, where heparin chemoprophylaxis is strongly recommended. Despite the signifi- The main limitation of our work is that it is a cant morbidity associated with thromboembolic retrospective study, where the patients were not ran- complications, Broughton and colleagues re- domized with respect to heparin chemoprophylaxis.
port inconsistent practice among plastic surgeons It is possible that the decision to use heparin can be in the implementation of deep vein thrombosis influenced by events of the operation; a surgeon may shy away from heparin use in cases where excessivebleeding was encountered during flap dissection.
The study is also subject to ascertainment bias, where CONCLUSIONS
postoperative examination may focus more on the Our study suggests that the use of heparin for TRAM flap on the chest than the abdominal donor venous thrombotic prophylaxis does not increase site. Therefore, a hematoma affecting the breast the risk of operative hematoma after TRAM flap mound contour may be detected more readily breast reconstruction. Our patient population than a hematoma developing in the abdominal serves as a good model for other procedures in site. This study would also not detect small hema- plastic surgery that involve wide soft-tissue under- tomas that did not warrant operative evacuation, mining, such as open excision lipectomy of the or clinically silent deep vein thrombosis and pul- torso. We are encouraged by these results and present this work as the basis for future investiga- One strength of this study is that the patient tion, such as validation of these findings with a population reviewed is representative of the typi- prospective, randomized, controlled trial. Evi- cal plastic surgery patient, middle aged and fe- dence-based practice of venous thrombotic che- male. Furthermore, breast reconstruction using moprophylaxis in plastic surgery is clearly impor- abdominal tissue provides a model of extensive tant, as our patients undergoing elective surgery soft-tissue undermining and dissection, which may are generally healthy and thromboembolic com- be applicable to other plastic surgery operations of the torso and abdomen that involve similar open Dennis P. Orgill, M.D., Ph.D.
dissection. The dissection of TRAM flaps in the context of immediate and delayed breast recon- struction involves elevation of a wide area of ab- dominal tissue, undermining of breast skin, and flap transfer (rotation or microvascular). An ad- ditional strength of this study is the scale, in termsof both the number of cases examined and the REFERENCES
number of surgeons (n ϭ 14) involved. The large 1. Geerts, W. H., Heit, J. A., Clagett, G. P., et al. Prevention of venous thromboembolism. Chest 119: 132S, 2001.
number of surgeons contributes to reduce any bias 2. Anderson, F. A., Jr., Wheeler, H. B., Goldberg, R. J., et al. A that may be related to surgical technique.
population-based perspective of the hospital incidence and Risk assessment and algorithms guiding the case-fatality rates of deep vein thrombosis and pulmonary use of heparin for thromboembolic chemopro- embolism: The Worcester DVT Study. Arch. Intern. Med. 151: phylaxis for patients undergoing plastic surgery 3. Dismuke, S. E., and Wagner, E. H. Pulmonary embolism as have been summarized previously.5,23 Davison and a cause of death: The changing mortality in hospitalized colleagues formulated a useful work sheet that patients. J.A.M.A. 255: 2039, 1986.
stratifies patient risk factors and recommends 4. Sakon, M., Maehara, Y., Yoshikawa, H., et al. Incidence of methods of thromboembolic prophylaxis. 23 Their venous thromboembolism following major abdominal sur- recommendations are based on evidence summa- gery: A multi-center, prospective epidemiological study inJapan. J. Thromb. Haemost. 4: 581, 2006.
rized from the American College of Chest Physi- 5. McDevitt, N. B. Deep vein thrombosis prophylaxis: American cians and existing risk-assessment models.20,35 Ac- Society of Plastic and Reconstructive Surgeons. Plast. Recon- cording to these sources, patients over the age of 60 undergoing any procedure involving general 6. Most, D., Kozlow, J., Heller, J., et al. Thromboembolism in anesthesia for greater than 1 hour are considered plastic surgery. Plast. Reconstr. Surg. 115: 20e, 2005.
7. Reinisch, J. F., Bresnick, S. D., Walker, J. W., et al. Deep high risk and should receive mechanical and hep- venous thrombosis and pulmonary embolus after face lift: A arin thromboembolic prophylaxis. Additional risk study of incidence and prophylaxis. Plast. Reconstr. Surg. 107: factors of malignancy (as in many reconstructive Volume 121, Number 4 • Hematoma Complication with Heparin 8. Rao, R. B., Ely, S. F., and Hoffman, R. S. Deaths related to 23. Davison, S. P., Venturi, M. L., Attinger, C. E., et al. Prevention liposuction. N. Engl. J. Med. 340: 1471, 1999.
of venous thromboembolism in the plastic surgery patient.
9. Grazer, F. M., and Goldwyn, R. M. Abdominoplasty assessed Plast. Reconstr. Surg. 114: 43E, 2004.
by survey, with emphasis on complications. Plast. Reconstr. 24. Durnig, P., and Jungwirth, W. Low-molecular-weight heparin and postoperative bleeding in rhytidectomy. Plast. Reconstr. 10. Matarasso, A., Swift, R. W., and Rankin, M. Abdominoplasty and abdominal contour surgery: A national plastic surgery 25. Broughton, G., II, Rios, J. L., Rohrich, R. J., et al. Deep venous survey. Plast. Reconstr. Surg. 117: 1797, 2006.
thrombosis prophylaxis practice and treatment strategies 11. Hester, T. R., Jr., Baird, W., Bostwick, J., III, et al. Abdomi- among plastic surgeons: Survey results. Plast. Reconstr. Surg. noplasty combined with other major surgical procedures: Safe or sorry? Plast. Reconstr. Surg. 83: 997, 1989.
26. Hartrampf, C. R., Scheflan, M., and Black, P. W. Breast re- 12. van Uchelen, J. H., Werker, P. M., and Kon, M. Complica- construction with a transverse abdominal island flap. Plast. tions of abdominoplasty in 86 patients. Plast. Reconstr. Surg. Reconstr. Surg. 69: 216, 1982.
27. Dupin, C. L., Allen, R. J., Glass, C. A., et al. The internal 13. Aly, A. S., Cram, A. E., Chao, M., et al. Belt lipectomy for mammary artery and vein as a recipient site for free-flap circumferential truncal excess: The University of Iowa ex-perience. Plast. Reconstr. Surg. 111: 398, 2003.
breast reconstruction: A report of 110 consecutive cases.
14. Clagett, G. P., Anderson, F. A., Jr., Geerts, W., et al. Preven- Plast. Reconstr. Surg. 98: 685, 1996.
tion of venous thromboembolism. Chest 114: 531S, 1998.
28. Schusterman, M. A., Kroll, S. S., Miller, M. J., et al. The free 15. Clagett, G. P., Anderson, F. A., Jr., Heit, J., et al. Prevention transverse rectus abdominis musculocutaneous flap for of venous thromboembolism. Chest 108: 312S, 1995.
breast reconstruction: One center’s experience with 211 con- 16. Clagett, G. P., Anderson, F. A., Jr., Levine, M. N., et al.
secutive cases. Ann. Plast. Surg. 32: 234, 1994.
Prevention of venous thromboembolism. Chest 102: 391S, 29. Kakkar, V. V., Cohen, A. T., Edmonson, R. A., et al. Low molecular weight versus standard heparin for prevention of 17. Clagett, G. P., and Reisch, J. S. Prevention of venous throm- venous thromboembolism after major abdominal surgery: boembolism in general surgical patients: Results of meta- The Thromboprophylaxis Collaborative Group. Lancet 341: analysis. Ann. Surg. 208: 227, 1988.
18. Bick, R. L., and Haas, S. K. International consensus recom- 30. Koch, A., Bouges, S., Ziegler, S., et al. Low molecular weight mendations: Summary statement and additional suggested heparin and unfractionated heparin in thrombosis prophy- guidelines. European Consensus Conference, November laxis after major surgical intervention: Update of previous 1991. American College of Chest Physicians consensus state- meta-analyses. Br. J. Surg. 84: 750, 1997.
ment of 1995. International Consensus Statement, 1997.
31. Nurmohamed, M. T., Rosendaal, F. R., Buller, H. R., et al.
Med. Clin. North Am. 82: 613, 1998.
Low-molecular-weight heparin versus standard heparin in 19. Lee, A. Y., Levine, M. N., Baker, R. I., et al. Low-molecular- general and orthopaedic surgery: A meta-analysis. Lancet 340: weight heparin versus a coumarin for the prevention of re- current venous thromboembolism in patients with cancer.
32. Leonardi, M. J., McGory, M. L., and Ko, C. Y. The rate of N. Engl. J. Med. 349: 146, 2003.
bleeding complications after pharmacologic deep venous 20. Geerts, W. H., Pineo, G. F., Heit, J. A, et al. Prevention of thrombosis prophylaxis: A systematic review of 33 random- venous thromboembolism: The Seventh ACCP Conference ized controlled trials. Arch. Surg. 141: 790, 2006.
on Antithrombotic and Thrombolytic Therapy. Chest 126: 33. Chang, D. W., Wang, B., Robb, G. L., et al. Effect of obesity 21. Spinal Cord Injury Thromboprophylaxis Investigators. Pre- on flap and donor-site complications in free transverse rectus vention of venous thromboembolism in the acute treatment abdominis myocutaneous flap breast reconstruction. Plast. phase after spinal cord injury: A randomized, multicenter Reconstr. Surg. 105: 1640, 2000.
trial comparing low-dose heparin plus intermittent pneu- 34. Wang, H. T., Hartzell, T., Olbrich, K. C., et al. Delay of matic compression with enoxaparin. J. Trauma 54: 1116, transverse rectus abdominis myocutaneous flap reconstruc- tion improves flap reliability in the obese patient. Plast. Re- 22. Leyvraz, P. F., Bachmann, F., Hoek, J., et al. Prevention of constr. Surg. 116: 613, 2005.
deep vein thrombosis after hip replacement: Randomised 35. Caprini, J. A., Arcelus, J. I., and Reyna, J. J. Effective risk comparison between unfractionated heparin and low mo- stratification of surgical and nonsurgical patients for venous lecular weight heparin. B.M.J. 303: 543, 1991.
thromboembolic disease. Semin. Hematol. 38: 12, 2001.

Source: http://www.portaldacirurgiaplastica.com.br/arquivos/30112009185424-arquivoa.pdf