Color duplex Doppler ultrasound scanning for detection of deep venous thrombosis in total knee and hip arthroplasty patients

The Journal of Arthroplasty Vol . 9 No . 5 1994

Color Duplex Doppler Ultrasound Scanning for Detection of Deep Venous Thrombosis in Total Knee and Hip Arthroplasty Patients Incidence, Location, and Diagnostic Accuracy Compared With Ascending Venography Darr W . Leutz, MD, and E . Shannon Stauffer, MD

Abstract : Ninety-nine patients undergoing total hip arthroplasty (THA) or total knee arthroplasty (TKA) were prospectively studied for deep venous thrombosis (DVT) of the lower extremity . Eighty-three hips in 42 THA patients and 107 knees in 57 TKA patients were studied with noninvasive color duplex Doppler ultrasound flow scanning and ascending venography 3-9 days after surgery . In all patients, surveillance studies were performed within 24 hours . The ultrasonographer and radiologist were blinded to the study . The incidence of DVT was significantly greater following TKA (61%) than THA (17%) (P < .001) . After TKA . significantly more DVT was found in the ipsilateral (32 of 57 patients) versus contralateral knees 1,3 of 50 patients), and more thrombi were located below the knee (34 of 107 knees) than above the knee (1 of 107 knees) (P < .001) . More thrombi were found in the contralateral limb after THA (5 of 41 patients vs 3 of 42 patients), but this was not significant . Two patients developed nonfatal pulmonary emboli . The sensitivity of color flow scanning was relatively poor initially (67% of above-knee patients and 57% of below-knee patients), but improved in the last 50 patients (100% of above-knee patients and 79% of below knee patients) . Negative predictive values exceeded 90% in the THA group and 87% in the TKA group . Positive predictive values ranged from 89 to 100% . Specificity was always above 96% . Longitudinal ultrasound scanning with the newer Quantum 2000 angiodynograph (lssaquah, WA) provided the best image quality . Direct imaging of the thrombus and lack of augmentation was diagnostic of deep venous thrombosis as opposed to compressibility . Color duplex Doppler ultrasound flow scanning effectively predicts the presence of DVT in the above-knee segment in patients undergoing THA or TKA . The sensitivity in locating below-knee thrombi has significantly increased to 79% and appears to be technician-dependent . Key words : arthroplasty, thromboembolism, deep venous thrombosis, ultrasonography, color duplex Doppler scanning .

The purpose of this study was to report the inci-

in total hip arthroplasty (THA) and total knee

dence and location of deep venous thrombosis (DVT)

arthroplasty (TKA) patients, as well as the diagnostic accuracy of color duplex Doppler ultrasonography

Southern Illinois University, School of Medicine . Department of Surgery . Division ofOrthopaedicSurgergandRehabilitation, Springfield, Illinois. From

(color flow scanning )CFSJ) in identifying DVT . The incidence of DVT in the United States is about

Reprint requests- Darr W Leon . MD, Southern Illinois University, School of Medicine, Department of Surgery, Division of Orthopaedic Surgery and Rehabilitation, PO Box 19230, Springfield, IL 62794-9230 .

250,000 per year, and more than 100,000 patients die annually from pulmonary embolism .' The incidence of DVT in the total joint population has been

543



544

The Journal of Arthroplasty Vol . 9 No . 5 October 1994

reported to be 10-84% despite prophylaxis, with a frequency of pulmonary embolism of up to 2% . 2 ' Long-term sequelae, namely recurrent DVT, venous insufficiency, and postphlebitic syndrome, have been shown to be disabling .' Clinical diagnosis is often unreliable and inconsistent .' Noninvasive, reliable, accurate, and repeatable detection of thrombi is needed . Ascending venography is accepted as the gold standard for the diagnosis of DVT but it is invasive, lacks repeatability, causes discomfort, and has morbidity, including postvenographic DVT and local ulceration .` ' o Several noninvasive techniques have been devised in the past. The accuracy of conventional duplex scanning without color enhancement is well documented in the literature and appears to be superior to that of other noninvasive methods ." - "However, conventional duplex scanning has limitations in the detection of DVT in the calf veins and adductor canal region . 15-19 Some studies have recently shown high false negative rates using conventional real-time Bmode ultrasound in the surveillance of patients after THA and caution against their use as the sole diagnostic modality in this high-risk population .-O" Color flow scanning is a recent improvement in the noninvasive detection of DVT (Fig . 1), but there are only a few reports documenting the accuracy of CFS in suspected DVT.' 2-J.5 Recently, we have shown the accuracy of CFS in the surveillance of DVT in postoperative asymptomatic patients versus the diagnostic accuracy in patients with symptomatic DVT .'° To our knowledge, no prospective studies have been performed regarding the accuracy of CFS in the diagnosis of DVT in total joint arthroplasty patients .

Fig . 1 . A normal color duplex Doppler scan of the posterior tibial artery (PT Art) and veins and the peroneal veins without thrombus .

Materials and Methods From July 1989 to February 1991, 99 patients (190 limbs) who underwent THA (42 patients, 83 limbs) or TKA (57 patients, 107 limbs) were included in the study . All patients gave informed consent, and the study was approved by the local human subject research committee . All patients underwent bilateral CFS and ascending venography 3-9 days after surgery . Surveillance studies were performed within 24 hours . Two patients who had bilateral total joint arthroplasty in separate admissions to the hospital were entered mto the study group twice . Eight patients refused contralateral ascending vcnography because of the pain they experienced from the ipsilateral venograrn, which was performed first . All venograms were performed using a modification of the technique described by Rabinov and Paulin27 and were interpreted independently by an experienced radiologist who was blinded to the study . Patients were excluded from the study if bilateral limb evaluations were inadequate, unsuccessful, or not performed secondary to patient refusal. A persistent intraluminal filling defect with venography was considered to be direct evidence for DVT . Color flow studies were initially performed with the QAD I angiodynograph (Quantum Medical Systems, Issaquah, WA) . Patients were scanned supine in the Trendelenhurg inclination with the lower extremities externally rotated . During the final 14 months of the study, CFS was performed with an updated model of the QAD I, the Quantum 2000 (Quanturn Medical Systems) . This newer version has an improved clarity of the image in the slow flow states . With CPS, stationary tissues are gray, while moving erythrocytes are color coded red or blue depending on the direction and magnitude . A total lack of color is indicative of the absence of flow or thrombus (Fig . 2) . All scans were interpreted by a member of the vascular surgical staff who was blinded to the results of the venograms . A scan was considered positive if thrombus was visualized by encroachment of flow image, absence of flow image, absence of flow despite augmentation, or noncompressibility of the venous segment . A negative scan was determined if no thrombus was found and augmented flow was normal . Equivocal scans were considered positive for DVT; if there was no identifiable thrombus, but spontaneous or augmented flow patterns were abnormal . Ascending venography was assumed to be the diagnostic standard for detection of DVT, with sensitivity and specificity of 100% each . Statistical analyses were compared by chi-square or Fisher's exact tests, with significance at P - .05 . All patients received a

Color Doppler Scanning

•

Leutz and Stauffer

545

Fig . 2 . (A) Color scan of the superficial femoral artery (SFA) and superlicial femoral vein (SFV) . Note the large, partially occluding thrombus in the superficial femoral vein . (B) Color duplex Doppler scan of the lower leg demonstrating the peroneal and posterior tibia) vessels . Note the large occluding thrombus in the peroneal vein . (C) Posterior tibial vein thrombus in an asymptomatic calf .

variety of DVT prophylaxis, including compression stockings, aspirin, arid warfarin (Coumadin, DuPont, Wilmington, DE) . Thrombi were classified as ipsilateral or contralateral and above knee or below knee . Above-knee segments included the common femoral, superficial femoral, and popliteal veins . Below-knee segments included the posterior tibial, anterior Iibial, and peroneal veins, and were designated tibioperoneal . Results Ninety-nine Patients (190 limbs) who underwent total joint arthroplasty were studied . Fifty-seven patients had TKA and 42 had THA . Both groups were

comparable with respect to age, sex, height, weight, and diagnosis (Table I)- The overall incidence of DVT for both groups was 42 of 99 patients (42%), and was significantly greater in the TKA group (35 of 57 patients, 61%) versus the THA group (7 of 42 patients, 17%) (P < .001) (Table 2) . Significantly more thrombi were located below the knee (34 of 107 knees) than above the knee (1 of 107 knees) (P < .001) in the TKA group . Thrombi occurred primarily in the calf veins of the ipsilateral limb after TKA (31 of 57 knees) . In the THA group, more thrombi occurred in the contralateral below knee (5 of 41 patients), but this was not significant (Table 3) . Most documented DVT were asymptomatic for both groups (P < _001) (Table 4) . Both groups were com-

546

The Journal of Arthroplasty Vol . 9 No . 5 October 1994 Table 1 . Comparison of the Study Groups THA

n M F

Average age (years) (range) Height (cm) (range) Smoker Past history of deep venous thrombosis Diagnosis Osteoarthritis (%) Revision (% Rheumatoid arthritis (%) Avascular necrosis (%)

Group

42 19 23 69 .7 (30-85)

TKA

Group

57 15 42 70 .3 (46-86)

169 .1 (152-185) 164.8 (147-195) 5 5 3 4

29 7 1 5

(69 .0) (16 .7) (2 .4) (11 .9)

48 2 4 0

(84.2) (3 .5) (7 .0) (0 .o)

parable . All patients received DVT compression boots and pharmacologic prophylaxis using Coumadin or aspirin . The incidence of DVT between patients who used these types of prophylaxis was not significant. Complications were minimal throughout the study . Wound infection occurred in three patients, wound hematoma in four patients, and two patients suffered a nonfatal pulmonary embolism . Overall, CFS was initially insensitive for detecting thrombus compared with venography (Table 5) . The sensitivity was poor initially (67% of above-knee patients and 56% of below-knee patients), but improved significantly in the last 50 patients (100% of above-knee patients and 79% of below-knee patients) (P < .00 1) (Table 6) . Positive predictive values were good, ranging from 88 to 100%, and negative predictive values exceeded 80-90% . Specificity always remained above 95% for all patients studied .

Table 2 . Incidence of Deep Venous Thrombosis Between the Study Groups Group n(%)

THA

Patients Limbs Above knee

7/42 (16 .7) 7/83 (8 .4) 2/83 (2 .4)

Discussion

Group n(%)

TKA

P

35/57 (61 .41 35/107 (32 .7) 1/107 (0 .9)

< .001 < .001

4/107 (31 .8)

< .001

(CF, Se Poi)

Below knee

5/83 (6 .2)

(TP) CF, common femoral vein; SF . superficial femoral vein ; POP, popliteaI vein ; TP . anterior tibia), posterior tibia), and pemucal veins .

Table 3 . Incidence of Deep Venous Thrombosis : ipsilateral versus Contralateral Limb Group n (%

THA

Above knee Ipsilateral Contralateral Below knee Ipsilateral Contralateral

Group n (%

TKA

2/42 (4 .8) 0/41 (0 .0)

1/57 (1 .8) 0/50 (0 .0)

1/42 (2 .4) 4/41 (9 .8)

31/57 (544) 3/50 (6 .0)

Table 4 . Incidence of Symptomatic versus Asymptomatic Deep Venous Thrombosis Group n (%)

THA

Symptomatic Asymptomatic P

0/42 (0 .0) 7/42 (16 .7) <001

Group n (%)

TKA

6'57 (10 .5) 29,57 )509) 4 .001

This study reports the incidence of DVT in total joint arthroplasty patients and indicates that more DVT occurs in the TKA group (61%) than in the THA group (17%), and usually in the below-knee segment of the ipsilateral operative limb . Our report shows a 2% incidence of symptomatic pulmonary embolism, both of which were nonfatal . These data are consistent with previous reports ." - " Routine ventilationperfusion scans were not performed on this patient population . Most of the DVTs in our study occurred in the calf veins and were asylnptomatic . To our knowledge, only one Study" shows asymptomatic calf vein thrombosis to be the cause of pulmonary embolism in this population . Pellegrini et al .'s" numbers are small (3 of 5 patients), but they do question the significance of these thrombi . The need to survey arid/ or treat calf vein DVT remains unsolved and the significance of finding an asymptomatic calf vein thrombus remains in dispute . Lotke et al .' and Moser and LeMoine 33 agree that symptomatic or asymptomatic calf vein thrombosis poses a low risk for pulmonary embolism . Our results also show that CFS can effectively document the presence of DVT in the above-knee segment . We have previously shown poor results for routine surveillance with CFS in patients who are at high risk for DVT .` 6 .34 .35 Our accuracy with CFS below the knee was initially poor and appeared to be technician-dependent. In this series of our last 50 patients, however, the accuracy of CFS improved significantly, with sensitivity at 79% and specificity at 95% . We believe that the following resulted in im-

Color Doppler Scanning

•

Leutz and Stauffer

547

Table 5 . Accuracy of Color Flow Scanning

Sensitivity Specificity Positive predictive value Negative predictive value

Limb

Above Knee

Below Knee

n (%)

n (%)

n (%)

24/42 (57 .1) 143/148 (96 .6) 24/27 (88 .9) 143/163 (87.7)

2/3 (66 .6) 188/188 (100) 2/2 (100) 187/188 (99 .5)

24/42 (57-1) 144/148 (97 .3) 24/27 (88 .9) 144/163 (88 .3)

proved accuracy of CFS in this series : (1) improved image quality with the newer Quantum 2000 angiodynograph instead of the QUAD 1, (2) utilization of longitudinal scanning to image longer venous segments with each image, (3) concentration on the clear visualization of the thrombus or lack of flow augmentation instead of incompressibility of the vessel, and (4) reduction of the "edema attenuation" of the scan by elevation of the limb prior to scanning . Color flow scanning is very cost-effective when compared to venography . The total cost of CFS at our institution is $165 .00 in the Radiology Department for both extremities and $360 .00 in the Vascular Laboratory . The total charges for unilateral venography approach $450 .00, depending on the type of contrast used, and $900 .00 if bilateral . Therefore, one can expect a savings of $285 .00-$735 .00 per patient with CPS . Also, CFS is well tolerated by the patient . Eight patients in our study group refused contralateral venography due to the pain they experienced with ipsilateral venography . Our accuracy above the knee, which includes the popliteal veins, is now 100% . We agree with Moser and LeMoine" and Lotke et al . 78 that isolated calf vein thrombus may not be clinically significant . We, therefore, believe that CFS can effectively be used as the sole diagnostic criterion for diagnosing and locating clinically significant thrombi in postoperative THA and TKA patients . We use CFS as the diagnostic modality of choice based on its accuracy, noninvasiveness, ease of use, repeatability, cost-effectiveness, and patient acceptance .

Table 6 . Accuracy : First Half versus Second Half Patients (1-49) Patients (50-99)

Limb Sensitivity Specificity Above knee Sensitivity Specificity Below knee Sensitivity Specificity

n (%)

n (%)

P

5/10 (25 .0) 77/78 (99 .0)

19/24 (792) 66/68 (97.1)

.006

0/1 (0 .0) 97/97 (100)

2/2 (100) 90/90 (100)

5/19 (26 .3) 7980 (98 .8)

19/24 (792) 66/68 (97.1)

.007

Acknowledgments The authors would like to thank the following : Dr . Mark Mattos and Dr . David Sumner at the Division of Vascular Surgery, Southern Illinois University, for performing and interpreting the vascular studies, the Radiology Department at St . John's Hospital, Springfield, IL, for their assistance in performing and reading the venographic studies, the Orthopaedic Clinical faculty members at Southern Illinois University for providing the patient population for this study, and Mary Garfield and Anita Weinhoeft for their skills in the preparation of the manuscript .

References L Atik M: Venous thrombosis disease . p . 1043 . In Moore WS (ed) : Vascular surgery . Grune and Stratton, Orlando, 1986 2 . Coventry MB, Nolan DR, Beckenbaugh RD : "Delayed" prophylactic antieoagulauon : a study of results and complications in 2,012 total hip arthroplasties- J Bone Joint Surg 55A : 1487, 1973 3 . Francis CW, Marder VJ, Evarts EM, Yaukookhodi S : Two-step warfarin therapy : prevention of postoperative venous thrombosis without excessive bleeding . JAMA 249 :374, 1983 4 . Hull RD, Raskob GE : Prophylaxis of venous thromboembolic disease following hip and knee surgery . J Bone Joint Surg 68A :146, 1986 5 . Lagerstedt CI, Olsson CG, Fagher [30 et al : Need for long tern anticoagulant treatment m symptomatic calf-vein thrombosis . Lancet 7 :515 . 1985 6 . Lotke PA, Ecker ML, Abass A, Berkowitz H : Indications for the treatment of deep venous thrombosis following total knee replacement . J Bone . Joint Surg 66A : 202, 1984 7 . Cranlcy JJ, Canos AJ, Sull WJ : The diagnosis of deep venous thrombosis : fallibility of clinical symptoms and signs, Arch Surg 111 :34, 1976 8 . Albrechtsson U, Olsson CG : Thrombotic side effects of lower-limb phlebography. Lancet 3 :723, 1976 9 . Bettmann MA, Paulin 5 : Leg phlebography : [he incidence, nature, and modification of undesirable side effects . Radiology 122 :101, 1977 10 . Bettmann MA, Salzman EW, Rosenthal D : Reduction



548

The Journal of Arthroplasty Vol . 9 No . 5 October 1994

of venous thrombosis complicating phlebography . Am I Radiol 134 :1169, 1980 I I . Coinerota AJ, Katz ML, Grccnwald LL et al : Venous duplex imaging : should it replace hemodynamic tests for deep venous thrombosis? J Vase Surg It : 54, 1990 12 . Patterson PB, Fowl RJ, Keller IK et al : The limitations of impedance plethysmography in the diagnosis of acute deep venous thrombosis . J Vase Surg 5 :725 . 1989 13 . Rosner NH, Doris PE : Diagnosis of femoral popliteal venous thrombosis : comparison of duplex sonography and plethysmography . Am J Radiol 150 .623, 1988 14 . SumnerUS : Diagnosis of deep venous thrombosis . p . 1520 . In Rutherford RB (ed) : Vascular surgery . 3rd ed . WB Saunders, Philadelphia, 1989 15 . Barnes R . Nix L, Barnes L et al : Perioperative asymptomatic venous thrombosis : role of duplex scanning versus venography . 1 Vasc Surg 9 :251, 1989 16 . Cronan i1, Dorfman GS, Siola FH et al : Deep vein thrombosis : ultrasound assessment using vein compression . Radiology 162 :191, 1987 17 . Killewich LA, Bedford GR, Beach KW, Strandness DE : Diagnosis of deep venous thrombosis : a prospective study comparing duplex scanning to contrast venography . Circulation 79 :810, 1989 18 . Lensing A WA, Prandoni P, Brandies D et al : Detection of deep vein thrombosis by real-time B-mode ultrasonography . New Engl J Med 20 :342, 1989 19 . Wright DJ, Shepard AD, McPharlin M, Earnst CB : Pitfalls in lower extremity venous duplex scanning . J Vase Surg 1) : 675, 1990 20 . Borris LC, Christiansen HM, Lassen MR et al : Comparison of real-time B-mode ultrasonography and bilateral ascending phlebography for detection of postoperative deep vein thrombosis following elective hip surgery . Thromb Hacmost 61 :363, 1989 21_ Borris LC, Christiansen HM, Lassen MR et al : Realtime B-mode ultrasonography in the diagnosis of postoperative deep vein thrombosis in nonsyniptomatic high risk patients . Bur J Vasc Surg 4 :473, 1990 22 . Foley WD, Middleton WD, Lawson TL et al : Color Doppler ultrasound imaging of lower extremity venous disacase . Am 3 Radiol 152 :371, 1989 23 . Persson AV, Jones ( : . Zide R, Jewell ER : Use of triplex scanner in diagnosis of deep vein thrombosis . Arch Surg 124:593, 1989

24 . Rose SC, Zwiebel WJ, Nelson BD et al : Symptomatic lower extremity deep venous thrombosis : accuracy, limitations, and role of color duplex flow imaging in diagnosis . Radiology 175 :639, 1990 25 . Schindler JM, Kaiser M, Gerber A et al : Cutout coded duplex sonography in suspected deep vein thrombosis of the leg . BMJ 301 :1369, 1990 26 . Mattos MA, Londrey GL, Leutz DW et al : Color flow duplex scanning for the surveillance and diagnosis of acute deep venous thrombosis, J Vasc Surg 15 :366, 1992 27 . Rabinov K, Paulin S : Roentgen diagnosis of venous thrombosis in the leg . Arch Surg 104 :134, 1972 28 . Lotke PA, Won ,,, RY, Ecker ML : Asymplomanc pulmonary embolism after total knee replacement . Orthop Trans 10 :490, 1986 29 . Nelson CL, Day L, Hoaglund FT, Lotke PA : Symposium : deep venous thrombosis and pulmonary embolism in fractures about the hip . Contemp Orthop 18 : 717 . 1989 30 . Paiemant GD, Beisaw N, i .otke PA et al : Advances in the prevention of venous thromboembolic disease after hip and knee surgery . Orthop Rev Supp 18(suppl) :1, 1989 31 . Stulberg BN, Insall JN, Williams GW, Ghelman B : Deep vein thrombosis following total knee replacements : an analysis of six hundred and thirty-eight anhroplasties. J Bone Joint Surg 66A : 194, 1984 32 . Pellegrini VD, Francis CW, Harris C, Tottersman 5 : Embolic complications of calf thrombosis following total hip arthroplasty . Presented at the 58th Annual Meeting of the American Academy of Orthopaedic Surgeons, Anaheim, CA, April 1991 33 . Moner KM, LeMoine JR : Is embolic risk conditioned by location of deep venous thrombosis? Ann Int Med 94 :439, 1981 34 . Mattos MA, Lmrdrey GL, Leutz DW ct aL Color flow duplex scanning for the surveillance and diagnosis of acute deep venous thrombosis . Presented at 45th Annual Meeting of the Society for Vascular Surgery, Boston . MA, June 1991 35 . Sumner DS, Londrey GL, Spadone DP et al : Study of deep venous thrombosis in high risk patients using color flow Doppler. p. 63 . In Bergan JJ, Yao JST (eds) : Venous disorders . WR Saunders . Philadelphia . 1991