- Email: [email protected]
Comparison of duplex scanning and contrast arteriography: A community hospital experience
Comparison of duplex scanning and contrast artcriography: A community hospital experience J o h n W. Kenagy, M.D., Longview, Wash. In this study the results of ultrasonic scanning combined with spectral analysis (duplex scanning) are compared with contrast arteriograms. From an initial experience of more than 1000 duplex examinations 78 patients were identified who had independently interpreted carotid arteriograms within 1 month of their duplex study. Duplex and arteriogram reports were then compared on 156 individual carotid arteries. This study demonstrated that duplex scanning can identify hemodynamically significant carotid lesions with a high degree of accuracy. These lesions of greater than 50% diameter reduction were recognized with a sensitivity and specificity of 95%, compared with contrast radiography. We were also able to distinguish carotid occlusion with an accuracy of 93% (10 of 11). These results independently confirm the findings of other investigators and demonstrate the validity of duplex scanning in a community hospital as an accurate noninvasive examination in cerebral vascular disease. (J VAse SuR~ 1985; 2:591-3.)
Cerebral vascular disease remains a major health problem in the United States; 500,000 patients suffered a stroke in 1978 and 176,000 of them died. 1 Despite aggressive medical management and rehabilitative efforts, 20% to 30% of stroke victims die, and 80% of the survivors do not return to their previous level of activity. 2 Clearly prevention is the key to treatment of this disease. Since approximately 60% to 70% of strokes are associated with thromboembolic events, a attention has been directed to the diagnosis of arteriosclerotic disease in the readily accessible extracranial vascular system. Because of the cost and morbidity of carotid angiography, 4'5 a number ofnoninvasive examinations have been developed to evaluate the carotid bifurcation. Many of these tests involve the indirect evaluation of the carotid artery via collateral circulation and have been criticized for a lack of specificity.6 Recently, a new generation of instruments has been developed that directly evaluates flow at the internal carotid artery, characterized as the duplex examination because of its combined use of the ultrasonic imaging and spectral analysis of carotid flow. 7This study was instituted to compare the accuracy of duplex scanning with From the PeripheralVascular Service, St. John's Hospital. Reprint requests: John W. Kenagy,M.D., ClinicalAssistantProfessorof Surgery,Universityof Washington,LongviewGeneral and Thoracic Surgery, Inc., 900 Fir St., Suite l-J, Longview, WA 98632.
contrast arteriography and to evaluate the efficacy of duplex scanning in a community hospital setting. MATERIAL AND METHODS All duplex examinations were done on a commercially available duplex scanner (MK450PV, Mark V Scanner, Advanced Technology Laboratories, Bellevue, Wash.). Technically, this machine is characterized by the use of a 5 M H z pulsed Doppler with a focal range of 2 to 6 cm ( _+0.5 mm), a focal point of 4 cm (°0.5 cm), a crystal diameter of 9.5 cm ( _+0.5 ram), and a sample volume size of 2 x 4 ram. Spectral analysis was done with the Spectrum Flow Analyzer (model 459, Advanced Technology Laboratories). All examinations were done by a single ultrasound technician and interpreted by the author. The methods of interpretation of spectra for this study were modifications of those described by Blackschear et al.7 Lesions of >50% diameter reduction met the following criteria: (1) peak systolic frequency >5 M H z and (2) spectral broadening widening the frequency band and at least partially obscuring the "clear area" beneath. All scans werc taped and qucstionable examinations were revicwcd by the author. The interpretations wcrc made without knowledge of thc patient's artcriographic evaluation. Standard intra-arterial radiography was performed on these patients at the discretion of their rcfcrring physician. The rcsults 591
Journal of VASCULAR SURGERY
592 Kenagy
Table I. Comparison of duplex scanning and arteriography ~*
Duplex scanning was also very accurate in iden:. tifying internal carotid artery occlusion. O f 11 arteries occluded by arteriography, 10 were correctly identified by duplex scanning (10/11 x 100 = 93%).
A rteriography Duplex scan
<50% Stenosis
>50% Stenosis
Occluded
< 5 0 % Stenosis > 5 0 % Stcnosis Occluded
57 5 1
5 91 0
0 0 10
~Diagnosis of > 5 0 % stenosis: sensitivity = 95%; specificity = 95%; diagnosis of occlusion: sensitivity = 100%; specificity = 99%.
Table II. Indications for duplex scanning 1. Evaluation of asymptomatic carotid bruits 2. FoUow-up known carotid lesions 3. Design appropriate treatment plan for patients with symptomatic cerebral vascular disease 4. Evaluation of patients with vague cerebral symptoms to rule out significant carotid artery disease 5. Screen patients at high risk for cerebral vascular disease
were interpreted by three experienced radiologists, independently, without knowledge of the duplex examinations. The study was done by retrospectively reviewing the formal duplex and angiographic interpretations on 78 consecutive patients who underwent carotid arteriography and duplex scanning within 1 month of each other. No patients were excluded. Previous experience in our laboratory suggested that, given present methodology, we could not differentiate less than 50% stenotic lesions accurately. We therefore classified these lesions as <50% diameter reduction, >50% diameter reduction, or occluded. Because of the documented variability of arteriographic interpretations, 8 all angiographic reports that did not concur with duplex findings were reevaluated in&pendently by a single radiologist (Robert Hill, M.D.) with the method of quantitating carotid stenosis described by Brown and Johnston. 9 By this standard the final results were obtained. RESULTS Seventy-eight patients met the criteria for inclusion in this study; therefore a total of 156 carotid arteries were available for evaluation. Following initial comparison and reevaluation as described earlier, duplex scanning demonstrated a high degree of correlation with arteriography as demonstrated in Table I.
DISCUSSION
Duplex scanning represents a new generation in noninvasive carotid artery evaluation. By this method the internal carotid artery can be visualized in real time and its flow evaluated by spectral analysis. The major advantages of this technique over other forms of noninvasive carotid evaluation are: 1. Direct analysis of internal carotid flow is possible without relying on the carotid collateral circulation. 2. The sample for spectral analysis can be obtained ~ from midstream vessel flow by direct visualization. Other authors in an academic setting have found duplex scanning to be accurate compared with angiography. 1° We have attempted to answer the same questions as a private practice of vascular diagnosis and surgery in a community hospital setting. Our duplex unit is placed in St. John's Hospital, a 188-bed facility located in Longyiew, Wash., and administered by the Peripheral Vascular Services Department of that hospital. The Director of this unit (the author of this article) has complete responsibility for this diagnostic service. The unit employs two fulltime technicians, one of whom, with an extensive background in ultrasonography, does all duplex scans. All scans and spectra are interpreted by the Director in close cooperation with the technician. To evaluate the accuracy of our duplex service, we sought to compare our results with contrast radiography, well aware of the potential pitfalls in the use of this "gold standard. ''n Contrast radiography at our institution is done by one of three experienced vascular radiologists. Our comparison was made by retrospectively reviewing the independently reported duplex scans and arteriograms on 78 consecutive patients. Apparent discrepancies between reports were resolved by having one radiologist independently review the arteriograms in questions. The final results were then tabulated, which demonstrated a 95 % sensitivity and specificity of duplex scanning compared with arteriography in our institution. It was our personal experience early in this study that accuracy was not high in grading lesions of less
Volume2 Number4 July 1985
' b a n 50% diameter reduction. This corresponds to the experience of others with similar equipment.I2 New technical advances may greatly aid in accuracy of diagnosis in this important area, 13 but we have obviated the problem by making sure that our other referring physicians realize the limitations of this technique, as presently practiced. For example, our stated indications for duplex scanning are noted in Table II. All of these indications are predicated, to a greater or lesser degree, on identifying highgrade carotid arte U stenosis. We emphasize to all our referring physicians that some situations demand contrast radiography, even if duplex scanning shows normal flow (e.g., transient cerebral ischemia in a good-risk patient). The technique, however, remains invaluable in screening patients for high-grade ste-,osis. We are presently evaluating the combination o f intravenous digital subtraction angiography and duplex scanning to refine our diagnostic techniques further. SUMMARY
We have presented a study demonstrating the accuracy of duplex scanning in a community hospital environment. We believe the prerequisites for a quality duplex service include: 1. A technician with experience in ultrasonography or indirect carotid evaluation and an interest in mastering this technique 2. An interpreter with extensive background in the diagnosis and management of cerebral vascular disease and an adequate knoWledge of the benefits and limitations of this technique 3. Close cooperation and communication between technician and interpreter 4. Sufficient volume to maintain skills and a willingness to compare results with contrast radiography Given these prerequisites a duplex scrvice can then provide a valuable force in the management of patients with symptomatic and asymptomatic cerebral vascular disease.
Duplex scanning and contrast arteriography 593
I thank Mr. Glen Craft and Kathy Stafford, R.N., for performing the technical studies, and Robert Hill, M.D., for reviewing arteriographic examinations. REFERENCES
1. Arteriosclerosis, 1981. National Institutes of Health Publication, June 1981; 80:234. 2. Gresham GE, Fitzpatrick TE, Wolf PA, McNamara PM, Kamiel WB, Dawbes TR. Residual disability in survivors of stroke. Framingham study. N Engl J Med 1975; 293:954-6. 3. Kannel MD. Council on Cerebral Vascular Disease--American Heart Association. Stroke 1971; 2:423. 4. Mani RL, Eisenborg RL, McDonald EJ, Polloch JA, Marri JR. Complications of catheter cerebral arteriography: Analysis of 5,000 procedures. I. Criteria and indications. Am J Roentgenol 1978; 131:861-5. 5. Kerber CW, Cromwell LD, Dryer BP, Bank WO. Cerebral ischemia. I. Current angiographic techniques, complications in safety. Am J Roentgenol 1978; 130:1097-1103. 6. Carson SN, BlalsdellTW. New techniques in the evaluation of cerebral vascular disease. West J Med 1979; 131:355-63. 7. BlackschearJr WM, Phillips DJ, Thiele BL, Hirsch JH, Chikos PL, Marinelli MR, Ward KJ, Ward BA, Strandness Jr DE. Detection of carotid occlusivedisease by ultrasonic imaging and pulse Doppler spectrum analysis. Surgery 1979; 86:698-706. 8. Langlois Y, Roederer GO, Chan A, Phillips DJ, Beach KW, Martur D, Chikos DM, Strandness Jr DE. Evaluating carotid artery disease, the coincidence between pulsed Doppler spectrum analysis and anglography. Ultrasound Med Biol (In press.) 9. Brown PM, Johnston KW. The difficultyof quantifying the severity of carotid stenosis. Surgery 1982; 92:473-88. 10. Breslau PS, Knox RA, Phillips DJ, Beach KW, Chikos PM~ Thiele BL, StrandnessJr DE. Ultrasonic duplex scanningwith spectral analysis in extracranial carotid artery disease: Comparison with contrast arteriography. Vasc Diag Therapy 1982; 3:17-22. 11. Thiele BL, Chikos PM, Strandness JrDE. Arteriography the gold standard? Presented at the Third International Cardiovascular Congress, Phoenix, Ariz, Feb 15-17, 1981. 12. Fell G, Phillips DJ, Chikos PM. Ultrasonic duplex scanning for disease of the internal carotid artery, Circulation 1981; 64:1191. 13. Green Jr FM, Beach K, Strandness Jr DE, Fell G, Phillips DJ. Computer based pattern recognition of carotid arterial disease using pulsed Doppler ultrasound. Ultrasomld Med Biol 1982; 8:161-76.
Cerebral vascular disease remains a major health problem in the United States; 500,000 patients suffered a stroke in 1978 and 176,000 of them died. 1 Despite aggressive medical management and rehabilitative efforts, 20% to 30% of stroke victims die, and 80% of the survivors do not return to their previous level of activity. 2 Clearly prevention is the key to treatment of this disease. Since approximately 60% to 70% of strokes are associated with thromboembolic events, a attention has been directed to the diagnosis of arteriosclerotic disease in the readily accessible extracranial vascular system. Because of the cost and morbidity of carotid angiography, 4'5 a number ofnoninvasive examinations have been developed to evaluate the carotid bifurcation. Many of these tests involve the indirect evaluation of the carotid artery via collateral circulation and have been criticized for a lack of specificity.6 Recently, a new generation of instruments has been developed that directly evaluates flow at the internal carotid artery, characterized as the duplex examination because of its combined use of the ultrasonic imaging and spectral analysis of carotid flow. 7This study was instituted to compare the accuracy of duplex scanning with From the PeripheralVascular Service, St. John's Hospital. Reprint requests: John W. Kenagy,M.D., ClinicalAssistantProfessorof Surgery,Universityof Washington,LongviewGeneral and Thoracic Surgery, Inc., 900 Fir St., Suite l-J, Longview, WA 98632.
contrast arteriography and to evaluate the efficacy of duplex scanning in a community hospital setting. MATERIAL AND METHODS All duplex examinations were done on a commercially available duplex scanner (MK450PV, Mark V Scanner, Advanced Technology Laboratories, Bellevue, Wash.). Technically, this machine is characterized by the use of a 5 M H z pulsed Doppler with a focal range of 2 to 6 cm ( _+0.5 mm), a focal point of 4 cm (°0.5 cm), a crystal diameter of 9.5 cm ( _+0.5 ram), and a sample volume size of 2 x 4 ram. Spectral analysis was done with the Spectrum Flow Analyzer (model 459, Advanced Technology Laboratories). All examinations were done by a single ultrasound technician and interpreted by the author. The methods of interpretation of spectra for this study were modifications of those described by Blackschear et al.7 Lesions of >50% diameter reduction met the following criteria: (1) peak systolic frequency >5 M H z and (2) spectral broadening widening the frequency band and at least partially obscuring the "clear area" beneath. All scans werc taped and qucstionable examinations were revicwcd by the author. The interpretations wcrc made without knowledge of thc patient's artcriographic evaluation. Standard intra-arterial radiography was performed on these patients at the discretion of their rcfcrring physician. The rcsults 591
Journal of VASCULAR SURGERY
592 Kenagy
Table I. Comparison of duplex scanning and arteriography ~*
Duplex scanning was also very accurate in iden:. tifying internal carotid artery occlusion. O f 11 arteries occluded by arteriography, 10 were correctly identified by duplex scanning (10/11 x 100 = 93%).
A rteriography Duplex scan
<50% Stenosis
>50% Stenosis
Occluded
< 5 0 % Stenosis > 5 0 % Stcnosis Occluded
57 5 1
5 91 0
0 0 10
~Diagnosis of > 5 0 % stenosis: sensitivity = 95%; specificity = 95%; diagnosis of occlusion: sensitivity = 100%; specificity = 99%.
Table II. Indications for duplex scanning 1. Evaluation of asymptomatic carotid bruits 2. FoUow-up known carotid lesions 3. Design appropriate treatment plan for patients with symptomatic cerebral vascular disease 4. Evaluation of patients with vague cerebral symptoms to rule out significant carotid artery disease 5. Screen patients at high risk for cerebral vascular disease
were interpreted by three experienced radiologists, independently, without knowledge of the duplex examinations. The study was done by retrospectively reviewing the formal duplex and angiographic interpretations on 78 consecutive patients who underwent carotid arteriography and duplex scanning within 1 month of each other. No patients were excluded. Previous experience in our laboratory suggested that, given present methodology, we could not differentiate less than 50% stenotic lesions accurately. We therefore classified these lesions as <50% diameter reduction, >50% diameter reduction, or occluded. Because of the documented variability of arteriographic interpretations, 8 all angiographic reports that did not concur with duplex findings were reevaluated in&pendently by a single radiologist (Robert Hill, M.D.) with the method of quantitating carotid stenosis described by Brown and Johnston. 9 By this standard the final results were obtained. RESULTS Seventy-eight patients met the criteria for inclusion in this study; therefore a total of 156 carotid arteries were available for evaluation. Following initial comparison and reevaluation as described earlier, duplex scanning demonstrated a high degree of correlation with arteriography as demonstrated in Table I.
DISCUSSION
Duplex scanning represents a new generation in noninvasive carotid artery evaluation. By this method the internal carotid artery can be visualized in real time and its flow evaluated by spectral analysis. The major advantages of this technique over other forms of noninvasive carotid evaluation are: 1. Direct analysis of internal carotid flow is possible without relying on the carotid collateral circulation. 2. The sample for spectral analysis can be obtained ~ from midstream vessel flow by direct visualization. Other authors in an academic setting have found duplex scanning to be accurate compared with angiography. 1° We have attempted to answer the same questions as a private practice of vascular diagnosis and surgery in a community hospital setting. Our duplex unit is placed in St. John's Hospital, a 188-bed facility located in Longyiew, Wash., and administered by the Peripheral Vascular Services Department of that hospital. The Director of this unit (the author of this article) has complete responsibility for this diagnostic service. The unit employs two fulltime technicians, one of whom, with an extensive background in ultrasonography, does all duplex scans. All scans and spectra are interpreted by the Director in close cooperation with the technician. To evaluate the accuracy of our duplex service, we sought to compare our results with contrast radiography, well aware of the potential pitfalls in the use of this "gold standard. ''n Contrast radiography at our institution is done by one of three experienced vascular radiologists. Our comparison was made by retrospectively reviewing the independently reported duplex scans and arteriograms on 78 consecutive patients. Apparent discrepancies between reports were resolved by having one radiologist independently review the arteriograms in questions. The final results were then tabulated, which demonstrated a 95 % sensitivity and specificity of duplex scanning compared with arteriography in our institution. It was our personal experience early in this study that accuracy was not high in grading lesions of less
Volume2 Number4 July 1985
' b a n 50% diameter reduction. This corresponds to the experience of others with similar equipment.I2 New technical advances may greatly aid in accuracy of diagnosis in this important area, 13 but we have obviated the problem by making sure that our other referring physicians realize the limitations of this technique, as presently practiced. For example, our stated indications for duplex scanning are noted in Table II. All of these indications are predicated, to a greater or lesser degree, on identifying highgrade carotid arte U stenosis. We emphasize to all our referring physicians that some situations demand contrast radiography, even if duplex scanning shows normal flow (e.g., transient cerebral ischemia in a good-risk patient). The technique, however, remains invaluable in screening patients for high-grade ste-,osis. We are presently evaluating the combination o f intravenous digital subtraction angiography and duplex scanning to refine our diagnostic techniques further. SUMMARY
We have presented a study demonstrating the accuracy of duplex scanning in a community hospital environment. We believe the prerequisites for a quality duplex service include: 1. A technician with experience in ultrasonography or indirect carotid evaluation and an interest in mastering this technique 2. An interpreter with extensive background in the diagnosis and management of cerebral vascular disease and an adequate knoWledge of the benefits and limitations of this technique 3. Close cooperation and communication between technician and interpreter 4. Sufficient volume to maintain skills and a willingness to compare results with contrast radiography Given these prerequisites a duplex scrvice can then provide a valuable force in the management of patients with symptomatic and asymptomatic cerebral vascular disease.
Duplex scanning and contrast arteriography 593
I thank Mr. Glen Craft and Kathy Stafford, R.N., for performing the technical studies, and Robert Hill, M.D., for reviewing arteriographic examinations. REFERENCES
1. Arteriosclerosis, 1981. National Institutes of Health Publication, June 1981; 80:234. 2. Gresham GE, Fitzpatrick TE, Wolf PA, McNamara PM, Kamiel WB, Dawbes TR. Residual disability in survivors of stroke. Framingham study. N Engl J Med 1975; 293:954-6. 3. Kannel MD. Council on Cerebral Vascular Disease--American Heart Association. Stroke 1971; 2:423. 4. Mani RL, Eisenborg RL, McDonald EJ, Polloch JA, Marri JR. Complications of catheter cerebral arteriography: Analysis of 5,000 procedures. I. Criteria and indications. Am J Roentgenol 1978; 131:861-5. 5. Kerber CW, Cromwell LD, Dryer BP, Bank WO. Cerebral ischemia. I. Current angiographic techniques, complications in safety. Am J Roentgenol 1978; 130:1097-1103. 6. Carson SN, BlalsdellTW. New techniques in the evaluation of cerebral vascular disease. West J Med 1979; 131:355-63. 7. BlackschearJr WM, Phillips DJ, Thiele BL, Hirsch JH, Chikos PL, Marinelli MR, Ward KJ, Ward BA, Strandness Jr DE. Detection of carotid occlusivedisease by ultrasonic imaging and pulse Doppler spectrum analysis. Surgery 1979; 86:698-706. 8. Langlois Y, Roederer GO, Chan A, Phillips DJ, Beach KW, Martur D, Chikos DM, Strandness Jr DE. Evaluating carotid artery disease, the coincidence between pulsed Doppler spectrum analysis and anglography. Ultrasound Med Biol (In press.) 9. Brown PM, Johnston KW. The difficultyof quantifying the severity of carotid stenosis. Surgery 1982; 92:473-88. 10. Breslau PS, Knox RA, Phillips DJ, Beach KW, Chikos PM~ Thiele BL, StrandnessJr DE. Ultrasonic duplex scanningwith spectral analysis in extracranial carotid artery disease: Comparison with contrast arteriography. Vasc Diag Therapy 1982; 3:17-22. 11. Thiele BL, Chikos PM, Strandness JrDE. Arteriography the gold standard? Presented at the Third International Cardiovascular Congress, Phoenix, Ariz, Feb 15-17, 1981. 12. Fell G, Phillips DJ, Chikos PM. Ultrasonic duplex scanning for disease of the internal carotid artery, Circulation 1981; 64:1191. 13. Green Jr FM, Beach K, Strandness Jr DE, Fell G, Phillips DJ. Computer based pattern recognition of carotid arterial disease using pulsed Doppler ultrasound. Ultrasomld Med Biol 1982; 8:161-76.