Deep Vein Thrombosis Among Patients Entering Cardiac Rehabilitation After Coronary Artery Bypass Surgery* Marco Ambrosetti, MD; Mario Salerno, MD; Mara Zambelli, MD; Filippo Mastropasqua, MD; Roberto Tramarin, MD; and Roberto F. E. Pedretti, MD Background: Little information is available about the prevalence of deep vein thrombosis (DVT) after discharge from cardiac surgery units and its impact on rehabilitation programs. Objectives: To estimate the rate of DVT, in relation to different thromboprophylaxis strategies, in patients with a recent coronary artery bypass graft (CABG) entering cardiac rehabilitation. Methods: Two hundred seventy consecutive patients admitted to three rehabilitation facilities after CABG surgery from 19 cardiac surgery units (male patients, 81%; mean SD age, 64 9 years; interval after operation, 4 to 19 days) underwent serial leg venous ultrasound examination on admission to three rehabilitative units. Results: At admission, antiplatelet treatment was present in all patients except 10 with absolute contraindications. In 171 patients (63%), heparin prophylaxis (low-molecular-weight heparin once daily, 87%; unfractionated heparin twice daily, 13%) was reported, limited to the early postoperative period (< 3 days) in 102 patients (38%). DVT was detected in 47 patients (17.4%). The rate of proximal and isolated distal DVT was 2.6% (7 cases) and 14.8% (40 cases), respectively. DVT was complicated in two cases (0.7%) by symptomatic pulmonary embolism, fatal in one case (0.4%). Clots were found in the leg contralateral to the saphenous vein harvest site in half of all DVT cases. Forty-three DVT cases (91%) were diagnosed at admission, while serial ultrasound testing allowed diagnosis of an additional 4 distal DVT cases. At multivariate analysis, female sex (p < 0.001) and length of stay in the surgery unit > 8 days (p < 0.05) were independently associated with risk of DVT in the rehabilitation setting. The adoption of heparin prophylaxis until discharge predicted the absence of DVT after adjustment for immobility (p < 0.05). Conclusions: This study showed a high rate of DVT in patients entering cardiac rehabilitation after CABG surgery. Wearing unilateral graded compression stockings after CABG surgery had limited efficacy, as clots were often localized in legs contralateral to the saphenous vein harvest site. (CHEST 2004; 125:191 196) Key words: cardiac rehabilitation; coronary bypass; deep vein thrombosis; prevention; prophylaxis Abbreviations: CABG coronary artery bypass graft; DVT deep vein thrombosis; GCS graded compression stockings; PE pulmonary embolism; VTE venous thromboembolism *From the Division of Cardiology, IRCCS Fondazione Salvatore Maugeri, Institutes of Tradate (Drs. Ambrosetti, Salerno, and Pedretti), Pavia (Drs. Zambelli and Tramarin), and Cassano delle Murge (Dr. Mastropasqua), Italy. Manuscript received January 14, 2003; revision accepted June 10, 2003. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: permissions@chestnet.org). Correspondence to: Marco Ambrosetti, MD, Division of Cardiology, IRCCS Fondazione Salvatore Maugeri, Via Roncaccio 16, I 21049 Tradate (VA), Italy; e-mail mambrosetti@fsm.it Venous thromboembolism (VTE) is a common complication after major surgeries, including general, orthopedic, gynecologic, trauma, and neurosurgery, despite specific prophylaxis recommendations. 1 Little has been published concerning the frequency of deep vein thrombosis (DVT) and pulmonary embolism (PE) after coronary artery bypass graft (CABG) surgery. The number of patients who enter supervised rehabilitation programs early after cardiac surgery (fast-track protocols) is constantly increasing. Unlike other medical and surgical problems (such as congestive heart failure, bypass occlusion, or wound infection), which clearly complicate the immediate postoperative course, DVT may remain occult and untreated throughout the stay in the surgery unit, www.chestjournal.org CHEST / 125 / 1/ JANUARY, 2004 191
thus constituting a potential comorbidity at admission to rehabilitation facilities. To our knowledge, data on the prevalence and characteristics of patients with DVT when entering a phase II rehabilitation program after CABG surgery have not been reported. Early ambulation, antiplatelet drugs, graded compression stockings (GCS), and intermittent pneumatic compression currently represent the most frequently adopted preventive measures against VTE after CABG surgery. Nevertheless, the optimal strategy for thromboprophylaxis and the role of heparin have not yet been definitely stated by the latest American College of Chest Physicians Consensus Conference on Antithrombotic Therapy. 1 The objective of this study was to estimate the rate of DVT in patients admitted to cardiac rehabilitation after recent CABG surgery, and its relation to different postoperative prophylaxis strategies. Study Characteristics Materials and Methods We performed a prospective observational trial. The primary end point was to calculate the number of DVT cases, as detected by ultrasonographic testing, among patients entering a cardiac rehabilitation program after CABG. Secondary end points were to find clinical variables predictive of DVT at admission to rehabilitative facilities and to calculate the number of clinically evident PE cases during the rehabilitation period. Inclusion Criteria All consecutive patients admitted to three units of the Department of Cardiology of the Fondazione Salvatore Maugeri (Tradate, Pavia, and Cassano delle Murge, Italy) for cardiac rehabilitation after CABG surgery from May 2000 through April 2001 were considered. Patients with a previous history of VTE disease, concomitant other cardiac surgery, or planned anticoagulation (ie, for chronic atrial fibrillation or mechanical prosthetic valves) were excluded. Each patient was visited at admission by a trained physician who obtained a clinical history and evaluated the patient s eligibility for the study. Prophylaxis strategies against VTE adopted after CABG surgery were evaluated by review of the patient s clinical record. Once admitted, patients were noninvasively screened for DVT using a serial testing methodology. Those with DVT received antithrombotic treatment according to published guidelines, 1 while those without DVT continued or were de novo prescribed heparin prophylaxis if considered at high risk for VTE disease. During the hospital stay, the adoption of GCS in all patients was ensured. The scientific board approved the study protocol. Detection of VTE Within 2 days after admission, patients underwent a first-leg venous ultrasound examination. Ultrasonography was performed by a pool of five experienced physicians using Acuson 128 or Sequoia ultrasound systems (Acuson Computed Sonography; Mountain View, CA) equipped with 4-MHz and 7.5-MHz transducers with duplex and color Doppler capability. Both legs were scanned from the groin distally, including calf veins. The common femoral, deep femoral, superficial femoral, and popliteal veins were imaged. 2 Imaging of the calf veins (anterior tibial, posterior tibial, peroneal, sural, soleus muscle, and draining veins of the lateral and medial heads of the gastrocnemius muscle) was obtained both transversely and longitudinally and confirmed with color-flow accentuation after the lower third of the calf was squeezed. Proximal DVT was defined as incomplete compression of common femoral, superficial femoral, deep femoral, or popliteal veins (with or without thrombosis of the calf vein). Isolated calf DVT was defined as loss of compressibility and loss of the expected increase in color Doppler signal with flow augmentation. If the finding of initial ultrasonography was normal or inadequate for interpretation, testing was repeated 7 days later using the same methodology. The second ultrasonic finding was recorded as normal or abnormal. The examination of calf veins was interpreted by two independent observers and the diagnosis of isolated calf DVT was made only when both observers were in agreement. Patients who had both initial and final testing as normal were closely monitored during in-hospital stay and further tested if specific symptoms (including leg pain, swelling, or tenderness) occurred. If the result of initial or repeated testing was positive for proximal or isolated calf DVT, anticoagulation was given unless contraindicated. Patients who presented symptoms and signs related to PE during the rehabilitation program underwent objective testing with lung scanning according to published guidelines. 3 Statistic Analysis Differences among continuous variables with normal distribution were examined by Student t test, while the Wilcoxon rank-sum test was used when continuous variables failed tests for normality. Differences among categorical variables were examined by 2 with Yates correction or Fisher exact tests where appropriate. A two-tailed p value 0.05 was considered statistically significant. Logistic regression analysis was performed in order to select independent predictors of DVT occurrence among those variables that showed a p value 0.1 at univariate analysis, with and without adjustment for immobility (a condition strongly suspected to underlie DVT, as in other surgical conditions) at admission to the rehabilitative unit. The SPSS package (SPSS; Chicago, IL) was used to perform statistical analysis. Results Of 339 patients considered for eligibility, 69 patients were excluded because of previous history of VTE (n 8, 12%), concomitant noncoronary surgery or valve replacement (n 43, 62%), and planned anticoagulation after CABG surgery for atrial fibrillation or other indications (n 18, 26%). Of the 270 eligible patients, all gave informed consent to participate and were studied according to the protocol. Patients included in the study (Table 1) were discharged from 19 cardiac surgery units and admit- 192 Clinical Investigations
Table 1 Characteristics of CABG Patients at Admission Variables Data (n 270)* Male gender 221 (82) Female gender 49 (18) Mean age, yr 65 9 Average body mass index 26.0 4.2 Cancer 13 (5) Documented PAOD 56 (21) Varicosity 32 (12) Documented congenital thrombophilia 0 LVEF 50% 64 (24) Cigarette smoking 65 (24) Hypertension 163 (60) Dyslipidosis 130 (48) Diabetes mellitus 68 (25) *Data are presented as No. (%) or mean SD. PAOD peripheral arterial obstructive disease; LVEF left ventricular ejection fraction. Including history of malignancy. Carotid plaques included. ted to the rehabilitation units 4 to 19 days after surgery (mean, 8 days). Of these, 246 patients (91%) had undergone saphenous vein harvest, bilaterally in 26 cases (10%). Prolonged ( 3 days) bed immobility or a length of stay in the cardiac surgery unit exceeding 8 days occurred in 35 patients (13%). Prophylaxis Strategies Two hundred sixty patients (96%) received an antiplatelet agent to maintain patency of the bypass graft. The following daily doses of antiplatelet agents were prescribed after surgery: aspirin, 100 mg, in 32% of patients; aspirin, 325 mg, in 25%; ticlopidine, 250 mg, in 23%; lysine acetylsalicylate, 160 mg, in 16%; and indobufen, 200 mg, in 4%. In 10 patients (4%), antiplatelets were not administered due to absolute contraindication. Frequency of the use of GCS and subcutaneous heparin in the surgical setting is shown in Table 2. Thirty-one patients (11%) received no heparin or Table 2 Heparin Prophylaxis and GCS in 270 Patients at Admission to Cardiac Rehabilitation Units After CABG Surgery Variables No. (%) No heparin 99 (37) Heparin used 3 d after CABG 102 (38) Heparin used 3 d after CABG 69 (25) Stopped before discharge from surgery unit 19 (7) Prescribed till discharge from surgery unit 50 (18) No GCS 70 (26) Ipsilateral GCS 182 (67) Bilateral GCS 18 (7) mechanical prophylaxis in the absence of a reported absolute contraindication; 27 of them (87%) were receiving antiplatelet medication. In 171 patients (63%), treatment with low, fixed doses of subcutaneous heparin was reported. Of these, 149 patients (87%) and 22 patients (13%) were prescribed lowmolecular-weight heparin once daily and 5,000 U bid of unfractionated heparin, respectively. Concerning low-molecular-weight heparin prophylaxis, all dosages adopted were effective, according to current recommendations, 1 in reducing the incidence of DVT after general surgery. Heparin prescription was limited to the early postoperative period 3 days after operation in 102 cases (38%). Patients who received heparin prophylaxis were not significantly different in age (65 10 years vs 65 9 years [mean SD], p 0.93), prevalence of postoperative complications (15% vs 10%, p 0.38), obesity (36% vs 30%, p 0.44), and cancer (5% vs 4%, p 0.85) from those who did not. GCS were prescribed postoperatively in 200 patients (74%), bilaterally in 18 patients (7%). In 28 patients (14%), delayed adoption ( 2 days) or interruption ( 3 h/d for 3 days) of GCS was reported. Frequency of VTE DVT was detected in 47 patients (17.4%), complicated in 2 patients (0.7%) by clinically evident PE. One death in the hospital (0.4%), the day after admission to the rehabilitative unit (seventh postoperative day), occurred. This 72-year-old woman received preventive heparin until discharge from the surgery unit and died from a massive PE 12 h after detection of isolated calf DVT, despite the fact that combined heparin and warfarin treatment was immediately administered. The other symptomatic PE occurred in a 66-year-old woman the day after anticoagulation was administered because of bilateral proximal DVT. This patient, in whom heparin prophylaxis was prescribed for 4 days after surgery, underwent urgent cardiopulmonary embolectomy due to hemodynamic instability, 4 with success. Globally, proximal DVT was detected in seven patients (2.6%), while isolated calf DVT accounted for the remaining 40 cases (14.8%). In 23 patients (49% of all DVT cases), clots were found in the leg contralateral to the saphenous vein harvest site. Concerning the timing of DVT diagnosis, 43 DVT cases (91%) were diagnosed at first testing, 3 cases (7%) at repeat testing 7 days after the first evaluation, and 1 case (2%) at a third ultrasonographic examination performed 19 days after admission due to the appearance of leg pain and swelling. All four late diagnoses regarded isolated calf DVT. www.chestjournal.org CHEST / 125 / 1/ JANUARY, 2004 193
The frequency of DVT according to the shortterm postoperative prophylaxis strategy (ie, no heparin, heparin limited to the early postoperative period, heparin until discharge from the surgery unit) is shown in Figure 1. Subjects who acquired DVT were significantly older and had significant postoperative complications (leading to prolonged immobility or delayed discharge) more frequently than those without DVT (Table 3). A significantly higher prevalence of DVT among female patients was also observed. The adoption of postoperative heparin prophylaxis did not significantly vary between patients with DVT and patients without DVT as a whole (55% vs 65%, p 0.27). Among patients who adopted postoperative heparin, the rate of DVT did not significantly vary between groups with different treatment duration (15% in the group with any duration, 17% with treatment 3 days after surgery, and 10% with treatment until dismissal from the surgery unit; p 0.05). After adjustment for immobility, a weak correlation between the adoption of heparin prophylaxis until discharge from the surgery unit and absence of DVT in the rehabilitation setting was noted, as the rate of DVT was 8% and 21% in patients with and without heparin, respectively (p 0.095). In a multivariate logistic regression analysis, controlling for all significant variables on univariate analysis, age ( 65 years, p 0.001), gender (female sex, p 0.01), and postoperative complications (p 0.01) predicted the development of DVT after CABG surgery. After excluding bedridden patients at admission, the adoption of heparin prophylaxis until discharge from the surgery unit predicted the absence of DVT in the rehabilitation setting (p 0.05). Figure 1. Frequency of DVT according to the adoption of postoperative heparin prophylaxis (no heparin, heparin limited to the early postoperative period, heparin until discharge from the surgery unit). Discussion Our findings agree with previous studies 5,6 showing a relatively high rate of DVT after CABG (approximately 17%), in most cases without clinically significant symptoms. In our series, 2 of the 47 patients with DVT acquired PE, fatal in 1 patient. Postoperative PE is relatively uncommon after CABG surgery, with a detection rate of approximately 3%. 5,6 However, patients with PE have a high mortality rate (nearly 20%), and, when a nonfatal event occurs, length of stay and hospital costs significantly increase. The rate of DVT after CABG surgery has been evaluated in two different studies, 7,8 conducted at the same institution, using predischarge venous ultrasound examination. Goldhaber et al 7 enrolled the larger number of patients and reported an approximately 22% DVT rate. Nevertheless, specific guidelines for thromboprophylaxis after CABG surgery are not currently available, and few studies have investigated which preventive strategies are adopted in cardiac surgical facilities. At the same time, the impact of DVT on rehabilitation programs, which involve CABG patients at an ever earlier stage after surgery, is still unknown. The low number of published studies on DVT and PE after CABG surgery emphasizes the scant attention that these possible complications have received. Current thromboprophylaxis approaches after CABG surgery are based on early passive and active mobilization, GCS, intermittent pneumatic compression, the routine use of antiplatelet therapy, and subcutaneous heparin in selected cases. Antiplatelet therapy should not strictly be considered a specific measure against DVT and PE. 1 Concerning mechanical prophylaxis, the percentage of patients in our study who were prescribed bilateral GCS after operation was 10%, while the majority of patients received GCS ipsilateral to the leg whose saphenous vein was harvested. This prescription is based on the general concept that saphenous vein harvesting, rather than cardiac surgery per se, represents the most important risk factor for leg venous thrombosis. According to our observation, however, ipsilateral GCS alone after CABG surgery had limited efficacy, as clots were often found also in the leg in which the saphenous vein was not harvested. In patients who received heparin after cardiac surgery, the prophylaxis was often withdrawn within 3 days after CABG surgery, which may provide inadequate protection against VTE, especially in patients with ongoing risk factors. In our study, DVT rates in the subgroup of patients not receiving heparin prophylaxis (total DVT, 21%; proximal DVT, 5%; isolated calf DVT, 194 Clinical Investigations
Table 3 Comparison of CABG Patients With and Without DVT* DVT Variables All DVT (n 47) Proximal DVT (n 7) Isolated Calf DVT (n 40) No DVT (n 223) p Value Female gender 17 (36) 2 (28) 15 (37) 32 (14) 0.0001 Age, yr 70 7 71 5 70 7 63 9 0.001 Cancer 2 (4) 1 (14) 1 (2.5) 11 (5) 0.84 LVEF 50% 7 (15) 0 (0) 7 (17) 57 (26) 0.17 Cigarette smoking 6 (13) 2 (28) 4 (10) 59 (26) 0.08 Hypertension 28 (60) 4 (57) 24 (60) 135 (61) 0.87 PAOD 16 (34) 4 (57) 12 (30) 40 (18) 0.05 Obesity 18 (38) 3 (43) 15 (37) 73 (33) 0.57 Varicosity 8 (17) 1 (14) 7 (17) 25 (11) 0.39 Postoperative complications 15 (32) 2 (28) 13 (32) 20 (9) 0.0001 Urgent CABG 2 (4) 0 (0) 2 (5) 11 (5) 0.84 Saphenous vein used for CABG graft 46 (98) 7 (100) 39 (97) 200 (90) 0.13 Conservative saphenous vein harvesting 5 (11) 1 (14) 4 (10) 21 (9) 0.85 Postoperative heparin prophylaxis 26 (55) 2 (28) 24 (60) 145 (65) 0.27 (any duration) Postoperative heparin prophylaxis 12 (26) 2 (28) 10 (25) 57 (26) 0.86 ( 3 d duration) Postoperative heparin prophylaxis 5 (11) 1 (14) 4 (10) 44 (20) 0.21 (until dismission from surgery unit) Adoption of GCS 36 (77) 5 (71) 31 (77) 170 (76) 0.89 No compliance to or delayed 5 (11) 2 (28) 3 (7) 22 (10) 0.96 adoption of GCS Immobility at admission to rehabilitative unit 1 (2) 1 (14) 0 (0) 2 (1) 0.95 *Data are presented as No. (%) or mean SD. LVEF left ventricular ejection fraction. See Table 1 for expansion of other abbreviations. Statistical differences were reported between the whole DVT group (proximal plus isolated calf DVT) and the group without DVT. Obesity was defined as a body mass index 27 for men and 26 for women. If postoperative immobility 3 d or in-hospital stay 8d. Limited cutaneous incision. Unilateral or bilateral. 16%) [Fig 1] were not significantly different from those (22%, 4%, and 18%, respectively) reported by Goldhaber et al 7 among CABG patients treated with GCS and aspirin, confirming that in the absence of specific prophylaxis the incidence of this complication may be significant. The efficacy of routine heparin prophylaxis in prevention of VTE after CABG and the need for its use in well-defined patient populations are matters of debate. Prospective, randomized studies should be implemented to clarify this point. At the same time, the safety of heparin in this clinical setting should be evaluated, not only on account of the increased risk of bleeding, but also because of the described effect of CABG surgery in activating platelets and inducing the release of platelet factor 4, which could increase the risk of heparin-induced thrombocytopenia 6 and subsequent thrombotic complications. A possible limitation of our study could be the use of only a noninvasive imaging modality, ultrasonography, to detect DVT cases. Ultrasonography is highly sensitive and specific for proximal venous thrombosis (thrombosis in the popliteal, femoral, and iliac veins) in patients suspected of having a first episode of DVT, 9,10 but it may be less sensitive in detecting asymptomatic postoperative DVT (particularly if thrombi are confined to the deep veins of the calf). Due to its possibly limited sensitivity in this latter case, all patients underwent serial testing to identify extension of thrombosis into more proximal segments. Similarly, considering that approximately 50% of patients with DVT could have clinically silent PE, 11 the real incidence of PE in our study population was reasonably underestimated and much higher than our reported value of 1%. Another possible limitation of our study is that results cannot be generalized to all patients undergoing CABG, as it evaluated only patients admitted for postoperative rehabilitation, not all consecutive patients undergoing CABG among the 19 cardiac surgery units. In conclusion, this study estimated the rate of DVT in patients with a recent CABG entering a cardiac rehabilitation program, in relation to different prophylaxis strategies adopted. We found the following: (1) the rate of DVT in the rehabilitation setting is relatively high, (2) the use of thrombopro- www.chestjournal.org CHEST / 125 / 1/ JANUARY, 2004 195
phylaxis after CABG surgery is heterogeneous, and (3) ipsilateral GCS, alone, in the leg in which the saphenous vein was harvested has limited efficacy in preventing DVT. Data are required to indicate the best prophylactic approach with respect to efficacy and safety. ACKNOWLEDGMENT: We thank Domenico Scrutinio, MD, who supervised the study at the Institute of Cassano delle Murge, Italy; and Samuel Z. Goldhaber, MD, from the Department of Medicine, Brigham and Women s Hospital, Harvard Medical School, Boston, MA, for constructive comments of the manuscript. References 1 Sixth ACCP Consensus Conference on Antithrombotic Therapy. Chest 2001; 119 (Suppl):1S 370S 2 Birdwell BG, Raskob GE, Whitsett TL, et al. The clinical validity of normal compression ultrasonography in outpatients suspected of having deep venous thrombosis. Ann Intern Med 1998; 128:1 7 3 Task Force on Pulmonary Embolism, European Society of Cardiology. Guidelines on diagnosis and management of acute pulmonary embolism. Eur Heart J 2000; 21:1301 1336 4 Aklog L, Williams CS, Byrne JG, et al. Acute pulmonary embolectomy: a contemporary approach. Circulation 2002; 105:1416 1419 5 Josa M, Siouffi SY, Silverman AB, et al. Pulmonary embolism after cardiac surgery. J Am Coll Cardiol 1993; 21:990 996 6 Shammas NW. Pulmonary embolus after coronary artery bypass surgery: a review of the literature. Clin Cardiol 2000; 23:637 644 7 Goldhaber SZ, Hirsch DR, MacDougall RC, et al. Prevention of venous thrombosis after coronary artery bypass surgery (a randomized trial comparing two mechanical prophylaxis strategies). Am J Cardiol 1995; 76:993 996 8 Reis SE, Polak JF, Hirsch DR, et al. Frequency of deep vein thrombosis in asymptomatic patients with coronary artery bypass grafts. Am Heart J 1991; 122:478 482 9 Kearon C, Julian JA, Newman T, et al. Noninvasive diagnosis of deep vein thrombosis. Ann Intern Med 1998; 128:663 677 10 American Thoracic Society. The diagnostic approach to acute venous thromboembolism. Am J Respir Crit Care Med 1999; 160:1043 1066 11 Meignan M, Rosso J, Gauthier H, et al. Systematic lung scans reveal a high frequency of silent pulmonary embolism in patients with proximal deep venous thrombosis. Arch Intern Med 2000; 160:145 146 196 Clinical Investigations