331 and parity) was similar in patients without hiatus hernia. Moreover the biliary lipid composition of gallbladder bile from the patients operated on for hiatus hernia was clearly different from that of the bile of patients without hiatus hernia-the cholesterol saturation of the bile from the former group was significantly higher than that of the bile from controls. The cause of this association is unclear. It is unlikely that gallstone disease per se can induce the appearance of hiatus hernia or vice versa. The most likely explanaage,
body-weight
with
or
tion is that common setiological factors predispose to both cholelithiasis and hiatus hernia. These factors could be of dietary origin.s z3 za In favour of this hypothesis there are the following arguments: both cholelithiasis25-27 and hiatus hernia 28 29 are rare or non-existent in African or Asian people, who live in rural conditions, but very common in developed areas like Europe and North America.5 6 30 There is plenty of evidence that cholesterol gallstones could be related to carbohydrate consumption-i.e., reduction in intake of unrefined cereals and an increase in intake of sugar (and animal fat).5 Some authors have suggested that a fibre-depleted diet could favour the formation of cholesterol gallstones by inducing changes in bile-salt and cholesterol metabolism.3O-32 They have shown that bran reduces the cholesterol saturation of the bile of gallstone patients. 33 On the other hand, a low-residue diet may be one of the factors predisposing to hiatus hernia. It has been suggested that the often repeated effort of straining at stool associated with constipation results in increased intracolonic and intra-abdominal pressures.2 28 The exaggerated contractions necessary to propel the low fxcal content through the large bowel are believed to be a fundamental cause of diverticular disease.34 35 The raised intra-abdominal pressures are probably an important cause of hiatus hernia ("push-up" theory). 28 It thus seems likely that hiatus hernia and gallstones are causally related to fibre-depleted diets. The same may be true for the association of these two conditions with diverticular disease of the colon, usually called Saint’s triad. 36 Requests for reprints should be addressed to J. P. C. Clinique Médicale A, Centre Hospitalier Universitaire, 1 Place V. Pauchet, 80030 Amiens Cedex, France.
REFERENCES
1. Muller, C. J. B. S. Afr. med. J. 1948, 22, 376. 2. Forster, J. J., Knutson, D. L.J. Am. med. Ass. 1958, 168, 257. 3. Palmer, E. D. Am. J. med. Sci. 1962, 244, 70. 4. Kaye, M. D., Kern, F. Lancet, 1971, i, 1228. 5. Heaton, K. W. Clins. Gastroent. 1973, 2, 67. 6. Stein, G. N., Finkelstein, A. Am.J. dig. Dis. 1960, 5, 77. 7. Cohen, S., Harris, L. D. New Engl.J. Med. 1971, 284, 1053. 8. Kreel, L. Clins Gastroent. 1973, 2, 185. 9. Admirand, W. H., Small, D. M.J. clin. Invest. 1968, 47, 1043. 10. Holzbach, R. T., Marsh, M., Olszewski, M., Holan, K. ibid. 1973, 52, 1467. 11. Bell, G. D., White, J., Dowling, R. H. in Bile Acids in Human Diseases (edited by P. Back and W. Gerok); p. 155. Stuttgart, 1972. 12. Talalay, P. Methods biochem. Anal. 1960, 8, 119. 13. Bartlett, G. R.J. biol. Chem. 1959, 234, 466. 14. Van Lerenberghe, J., Cassaigne, R. Revue fr. etud. clin. biol. 1968, 13, 541. 15. Thomas, P. J., Hofmann, A. F. Gastroenterology, 1971, 65, 698. 16. Hegardt, F. G., Dam, H. Z. ErnährWiss. 1971, 10, 223. 17. Vogt-Moykopf, I., Dietz, R., Zeidler, D. Fortschr. Med. 1972, 90, 496. 18. Schmidt, H. D. Therapiewoche, 1974, 90, 496. 19. Block, M. A., Allen, H. M. Surg. Gynec. Obstet. 1971, 132, 46. 20. Friedman, G. D. Ann. intern. Med., 1968, 68, 222. 21. Friedman, G. D. J. chron. Dis. 1966, 19, 273.
VASCULAR-LABORATORY DIAGNOSIS OF CLINICALLY SUSPECTED ACUTE DEEP-VEIN THROMBOSIS D. PRESTON FLANIGAN STEVEN J. BURNHAM
JAMES J. GOODREAU JOHN J. BERGAN
JAMES S. T. YAO Blood Flow Laboratory and Division of Vascular Surgery, Department of Surgery, Northwestern University Medical School, Chicago, Illinois 60611, U.S.A.
Doppler ultrasound, impedance plethysmography, and contrast venography were performed in 207 lower limbs suspected of harbouring deep-venous thrombosis, to clarify the diagnos-
Summary
tic value and limitations of the non-invasive methods. Doppler ultrasound and impedance plethysmography were accurate in 96% and 95% of normal limbs, respectively. In limbs with venographic evidence of thrombosis requiring treatment, Doppler ultrasound and impedance plethysmography correctly detected thrombosis in 60% and 97%, respectively. Doppler ultrasound was 97% accurate in recognising chronic venous insufficiency. Impedance plethysmography was incorrectly positive in 74% of limbs with chronic venous insufficiency which had no venographically detected thrombosis. These findings suggest that, for the accurate diagnosis of clinically suspected deep-vein thrombosis, venography is necessary only in patients with chronic venous insufficiency who have normal Doppler ultrasound tests and abnormal impedance plethysmograms and in patients with abnormal cardiac hæmodynamics. In this series, 86% of limbs would have been spared venography had non-invasive tests been used. Venography, however, remains the standard test for the detection of minor calf-vein thrombosis. A diagnostic and therapeutic schema is proposed. Introduction NON-INVASIVE tests have lately been added to the vascular surgeon’s armamentarium for the diagnosis of venous disease. In a vascular laboratory these tests can be done by technicians; the results are interpreted by a physician who prepares a report to help the primary physician decide whether diagnostic venography or
therapeutic anticoagulation are required. The blood-flow laboratory of the Northwestern versity
McGaw Medical Center
uses a
Unicombination of
22. Donaldson, R. M. Gastroenterology, 1975, 68 1608. 23. Burkitt, D. P., Painter, N. S.J. Am. med. Ass. 1974, 229, 1068. 24. Mendeloff, A.I. New Engl J. Med. 1977, 297, 811. 25. Parnis, R. O. Trans. R. Soc. trop. Med. Hyg. 1964, 58, 437. 26. Shaper, A. G., Patel, K. M. East Afr. med. J. 1964, 41, 246. 27. Biss, K., Ho, K. J., Mikkelson, B., Lewis, L., Taylor, C. B. New Engl. J. Med. 1971, 284, 694. 28. Burkitt, D. P., James, P. A. Lancet, 1973, ii, 128. 29. Bassey, O. O., Eyo, E. E., Alinhanmi, G. A. Thorax, 1977, 32, 356. 30. Dyer, N. H., Pridie, R. B. Gut, 1968, 9, 696. 31. Pomare, E. W., Heaton, K. W. Br. med. J. 1973, iv, 262. 32. Low-Beer, T. S., Pomare, E. W. ibid. 1975, i, 438. 33. Pomare, E. W., Heaton, K. W., Low-Beer, T. S., Espiner, H. J. Am. J. dig. Dis. 1976, 21, 521. 34. Painter, N. S. Proc. R. Soc. Med. 1970, 63, 144. 35. Painter, N. S., Burkitt, D. P. Br. med.J.1971, 2, 450. 36. Burkitt, D. P., Walker, A. R. P.S. Afr. med. J. 1976, 50, 2136.
332
Doppler ultrasound and impedance plethysmography for non-invasive diagnosis of deep-venous thrombosis (D.V.T.). In an attempt to improve diagnostic accuracy in venous disease, Doppler ultrasound was used to detect flow reversal in femoral and popliteal veins, a common feature of patients with chronic venous insufficiency,’1 and to detect abnormal venous hsemodynamics due to cardiac insufficiency.Both conditions can yield abnormal impedance plethysmography results.2 We have evaluated the diagnostic accuracy of the non-invasive tests in comparison with contrast venography in patients with clinically suspected acute D.V.T. Correlation between the non-invasive tests and venograms was tabulated in seven specific groups-limbs with normal venograms and those with iliofemoral thrombosis, femoropopliteal thrombosis, major calf-vein thrombosis (two or more major veins occluded), minor calf-vein thrombosis (one major vein occluded), nonocclusive venous thrombosis, and chronic venous insufficiency. The results were then analysed to decide in which group, if any, venography might be eliminated and therapeutic decisions made on the basis of non-invasive testing alone.
capacitance seen in 3 s. If abnormal results were obtained, subsequent examinations were done after repositioning of the leg to check for positional venous obstruction. Tests were done before and after a 2 min exercise period consisting of dorsiflexion and plantar-flexion of the foot. The test was considered normal if post-exercise data fell into the normal range, even if pre-exercise data were abnormal. All tests were done by a trained technician, and reports were made for subsequent interpretation by a physician without the knowledge of symptoms and signs of the patient. venous
Results Normal Doppler ultrasound examination was determined by the presence of respiratory modulations in venous velocity waveforms, normal augmented signals, and Valsalva manoeuvre. Impedance plethysmography results were based on the criteria reported by Wheeler et ap,5 (fig. 1).
Materials and Methods 207 limbs were studied by Doppler ultrasound and impedance plethysmography in patients referred to the blood-flow laboratory because of suspected D.v.T. All patients underwent standard contrast ascending venography.
Doppler Ultrasound Doppler ultrasound examination
was done by strip-chart of directional velocity patterns from the common femoral and popliteal veins with a transcutaneous directional Doppler (Parks no. 806C). Changes in venous velocity in response to Valsalva and sound augmentation manoeuvre were done as previously described. 34
recordings
Impedance Plethysmography Impedance plethysmography was performed as described by Wheeler et al. Electrode cuffs were placed on the calf and a venous compression cuff was placed on the thigh. The hip and knee were flexed 30 degrees, the hip externally rotated, and the extremity elevated 20-30 degrees on pillows. The thigh cuff 43 cm water for 45 s, thus causing a rise in capacitance. After 45 s, the cuff was rapidly deflated and the slope of the fall in venous capacitance observed, The rise in venous capacitance was then plotted against the fall in was
inflated
to
venous
Fig. 1-Nomogram for impedance plethysmography. Based on the one devised by Dr H. B. Wheelers and confirmed by 8 Hull et al.
207 limbs studied, 98 were normal by contrast venography (table). 4 of these 98 limbs showed abnormal Doppler tests, giving a specificity of 96% (94/98) and a false-positive rate of 4%. In this report, "abnormal Doppler" refers to an examination showing evidence of D.V.T. An examination showing only flow reversal suggesting chronic venous insufficiency is referred to Of
EVALUATION OF NON-INVASIVE TESTING
I.P.G.
=
Impedance plethysmography. * Limbs without D.v.T.
(207 LIMBS)
333 such. 5 of the 98 normal limbs yielded abnormal impedance examinations for a specificity of 95% (93/98) and a 5% false-positive rate. All 18 limbs having iliofemoral D.V.T. on venography also had abnormal Doppler and impedance tests, thus yielding a 100% correlation in this group. Of 28 limbs with femoropopliteal thrombosis all showed abnormal impedance studies, but only 16 (57%) showed abnormal as
Doppler findings. showing major calf-vein thrombosis (thrombosis involving two or more segments), all 11 had accompanying abnormal impedance studies while only 4 of the 11 limbs (36%) showed abnormal Doppler examOf 11 venograms
,
inations. Of the 6 limbs with minor calf-vein thrombosis (thrombosis involving one segment or seen in soleal or gastrocnemius veins) demonstrated on venography, none showed abnormal Doppler tests while 1 had an abnormal impedance examination. There were 8 limbs where non-occlusive thrombosis (floating clot) was seen in. the popliteal veins on venography. Only 1 (12%) had an abnormal Doppler test and 6 (75%) had abnormal impedance examinations. 38 limbs demonstrated chronic venous insufficiency without thrombi seen on venography. 97% of these limbs had flow reversal on Doppler examination and 74% (28 limbs) demonstrated a falsely abnormal impedance examination. Discussion Various non-invasive
techniques such as phleborheography,6 strain-gauge plethysmography,7 Doppler ultrasound,3,4 and cuff impedance plethysmography2,1,1 have been used in the diagnosis of venous thrombosis. We chose Doppler ultrasound and impedance plethysmography because both methods are technically simple and can easily be done by technicians after a short training period. Although impedance plethysmography is more sensitive than Doppler ultrasound, it is less accurate in localising the site of thrombosis. Since impedance plethysmography had often been falsely positive in chronic venous insufficiency and cardiac insufficiency in our early experience, Doppler ultrasound was added in an attempt to detect these conditions. The sensitivity and specificity of the impedance plethysmography examination in this series were not significantly different from those reported by Wheeler and others.2 5 SAlthough Doppler ultrasound was excellent in the detection of iliofemoral thrombosis, the results in femoropopliteal and major calf-vein thrombosis were worse than those recorded by others. Patency of the profunda femoris vein has been thought in some instances to mask the detection of femoropopliteal thrombosis by Doppler ultrasound. Detection of major calf-vein thrombosis with Doppler ultrasound was primarily determined
by augmentation tests. Although only 36% of limbs with major calf-vein thrombosis were detected in this series by Doppler ultrasound, 100% were detected by impedance plethysmography. Barnes et al. have reported significantly better results in detecting calf-vein thrombosis with Doppler ultrasound. Our results with Doppler ultrasound of the calf are more in accord with those of Richards et al. 10 who also used the combination of impedance plethysmography and Doppler ultrasound. The discrepancy may be related to the fact that tests were
done by technicians with subsequent physician interpretation without knowledge of clinical signs and symptoms. Subjective interpretation of various diagnostic ultrasound manoeuvres by a physician performing the tests may increase the accuracy of interpretation. The high frequency of positive impedance tests in limbs with calf-vein thrombosis that we found contrasts with Wheeler’s latest study," where impedance tests were positive in only 6 of 27 limbs with calf-vein thrombosis. Impedance testing is of limited value when early thrombi fail to impede venous outflow or when there is duplication of the femoral vein.3 Our patients had clinically suspected D.V.T. of more than one day’s duration when tested in the vascular laboratory, so most of them had propagating thrombi large enough to give a positive impedance result. The false-positive rates were 3% for Doppler ultrasound and 24% for impedance plethysmography (table). If patients with chronic venous insufficiency and minor calf-vein thrombosis are excluded, impedance plethysmography yields only a 3% false-negative rate (all in the non-occlusive group), while Doppler ultrasound yields a 40% false-negative rate. Exclusion of patients with chronic venous insufficiency also changes the frequency of false-positive impedance plethysmography studies from 24% to 5% and of Doppler ultrasound from 3% to
4%. These results suggest a clinical schema for determining the need for venography and/or anticoagulation (fig.
2):
2-Clinical schema for patients with thrombosis.
Fig.
suspected acute venous
presence or absence of chronic venous insufficiency abnormal cardiac hsemodynamics should be determined by Doppler ultrasound. Patients with abnormal cardiac hasmodynamics should proceed to venography since D.v.T. in these patients cannot be diagnosed accurately by non-invasive methods. (2) When D.v.T. is suspected clinically but both Doppler ultrasound and impedance plethysmography are normal, limbs need not be tested by venography: only 3% of these limbs will have D.v.T. which should be treated. (3) If both non-invasive tests are abnormal venography is not necessary and the patient should be treated; in this way only 5% would be treated unnecessarily. (4) Limbs with abnormal Doppler ultrasound and normal impedance plethysmography were not encountered in our series.
(1) The
or
334
(5) Non-postphlebitic limbs with normal Doppler ultrasound impedance plethysmography need not be investigated by venography and should be treated; if this is done, only 5 % will be treated unnecessarily. (6) A clinical suspicion of D.V.T. superimposed on chronic venous insufficiency is more difficult to manage. When chronic venous insufficiency is detected by Doppler, an abnormal impedance examination is of no value (74% false-positive). Thus, only patients with postphlebitic limbs normal by Doppler and abnormal by impedance testing, and patients with abnormal cardiac haemodynamics, need proceed to venography for accurate diagnosis OfD.V.T. and abnormal
This plan would allow 3% of limbs with D.V.T. to go untreated and condemn 5% of limbs without D.V.T. to anticoagulant treatment and its inherent risks. Counterbalancing these drawbacks, however, are five advantages of avoiding venography,-the significant patient discomfort, radiation exposure, and the possibility of allergic reactions to contrast medium; the risk of induc-
ing phlebitis or worsening existing phlebitis by venography (as high as 20% in one series’2); non-invasive testing is about one-third the price of venography and does not require a doctor’s time other than for interpretation ; and contrast venography does not show well the
CONTROLLED CLINICAL TRIAL OF FIVE SHORT-COURSE (4-MONTH) CHEMOTHERAPY REGIMENS IN PULMONARY TUBERCULOSIS First Report of 4th EAST AFRICAN
AND
Five 4-mo regimens of chemotherapy for tuberculosis are compared. The two regimens in which rifampicin was given throughout the 4 mo were associated with bacteriological-relapse rates of 8% in the first 6 mo after stopping chemotherapy, but the three regimens in which rifampicin was given for only the first 2 mo had relapse-rates of 24-32%. There was no evidence that the addition of pyrazinamide in the second 2 mo of chemotherapy reduced the bacteriological-relapse rate. Removal of the streptomycin from the first 2 mo appeared to reduce the bactericidal and sterilising activity of the regimen, although the differences were not statistically significant. The incidence of adverse reactions was very low with all five regimens.
Summary
external and common iliac veins in 18% of cases.8 11t In
experience
laboratory approach
to guidewith patients diagnostic clinically susIn acute D.v.T. is our 179 of 207 series, possible. pected limbs (86%) could have been spared venography. 2 limbs with D.V.T. requiring treatment would have gone undeour
a
evaluation in
tected, and 5 normal limbs would have been treated if selection criteria had been used. Contrast venography can be deleted from the diagnostic evaluation of D.V.T. only when the alternatives are similar in sensitivity and specificity, cost less, and/or carry fewer risks and complications. For epidemiological or natural-history studies of acute D.v.T., non-invasive testing, which is flow-dependent, will not detect fresh thrombus or thrombi in the soleal sinuses or calf veins, or even major veins if the thrombus is not large enough to impede venous flow. Venography or 125I-fibrinogen remain the best techniques to establish the diagnosis of these conditions. our
Supported in part by the Dr Scholl Foundation University Vascular Research Fund.
and the Northwes-
tern
Requests for reprints should be addressed to J.S.T.Y., Ward Memorial Building, Medical School, 303 E, Chicago Avenue, Chicago, Illinois 60611, U.S.A.
REFERENCES
Alexander, R. H., Nippa, J. H., Folse, R. Am. Heart J. 1971, 82, 86. Wheeler, H. B., Mullick, S. C., Anderson, J. N., Pearson, D. Surgery, 1971, 70, 20. 3. Barnes, R. W., Russell, H. E., Wu, K. K., Hoak, J. D. Surg. Gynec. Obstet. 1976, 143, 425. 4. Yao, J. S. T., Gourmos, C., Hobbs, J. T. Lancet, 1972, i, 1. 5. Wheeler, H. B., O’Donnell, J. R., Anderson, F. A. Prog. cardiovasc. Dis. 1974, 17, 199. 6. Cranley, J. J., et al. Surg. Gynec. Obstet. 1975, 141, 331. 7. Barnes, R. W., Hokanson, D. E., Wu, K. K., Hoak, J. D. Surgery, 1977, 82, 1. 2.
219. Hull, R. W., et al. Circulation, 1976, 53, 696. Hull, R., et al. New Engl.J. Med. 1977, 296, 1497. Richards, K. L., et al. Archs intern. Med. 1976, 136, 1091. Benedict, K. T., Wheeler, H. B., Patwardhan, N. A. Radiology, 696. 12. Albrechtsson, U., Olsson, C. G. Lancet, 1976, i, 723.
8. 9. 10. 11.
Study
BRITISH MEDICAL RESEARCH COUNCILS
Introduction THE high level of efficacy of several 6 mo regimens of chemotherapy for pulmonary tuberculosis has been established in studies in East Africa and in many other parts of the world, relapse-rates after the end of chemotherapy being 5% or less. 1-11It was important, therefore, to investigate regimens of even shorter duration. A pilot study in France12 of two 3-mo regimens of rifampicin, isoniazid, and streptomycin had had a relatively low overall relapse-rate of 13%. Studies in East Africa5-7 had also shown that an intensive initial four-drug phase with streptomycin, isoniazid, rifampicin, and pyrazinamide given daily for 2 mo is a very effective component of short-course chemotherapy and that the pyrazinamide plays an important part, as has been confirmed in Hong Kong.8 The present study was planned to investigate whether adding pyrazinamide to the three drugs studied in France and also increasing the total duration of chemotherapy to 4 mo would improve the results. Even if not uniformly successful, this regimen might still be applicable in developing countries, where standardduration chemotherapy with regimens known to be 100% effective in antibacterial terms produces relatively low levels of success under programme conditions, 13 default from treatment being a major reason for failure. Thus, a 4 mo regimen that is . less than uniformly successful might still constitute a major therapeutic advance. Because daily administration of streptomycin in many developing countries almost invariably requires hospital admission for the duration of treatment, a regimen not including streptomycin was also investigated. This regimen, as well as being applicable to outpatients, provides direct information about the role of streptomycin in the initial intensive phase. Patients and Methods PLAN AND CONDUCT OF STUDY
Selection of 1977, 125,
Patients
-
Patients selected were Africans aged 15-65, with previously untreated extensive pulmonary tuberculosis believed to be of
body-weight
with
or
tion is that common setiological factors predispose to both cholelithiasis and hiatus hernia. These factors could be of dietary origin.s z3 za In favour of this hypothesis there are the following arguments: both cholelithiasis25-27 and hiatus hernia 28 29 are rare or non-existent in African or Asian people, who live in rural conditions, but very common in developed areas like Europe and North America.5 6 30 There is plenty of evidence that cholesterol gallstones could be related to carbohydrate consumption-i.e., reduction in intake of unrefined cereals and an increase in intake of sugar (and animal fat).5 Some authors have suggested that a fibre-depleted diet could favour the formation of cholesterol gallstones by inducing changes in bile-salt and cholesterol metabolism.3O-32 They have shown that bran reduces the cholesterol saturation of the bile of gallstone patients. 33 On the other hand, a low-residue diet may be one of the factors predisposing to hiatus hernia. It has been suggested that the often repeated effort of straining at stool associated with constipation results in increased intracolonic and intra-abdominal pressures.2 28 The exaggerated contractions necessary to propel the low fxcal content through the large bowel are believed to be a fundamental cause of diverticular disease.34 35 The raised intra-abdominal pressures are probably an important cause of hiatus hernia ("push-up" theory). 28 It thus seems likely that hiatus hernia and gallstones are causally related to fibre-depleted diets. The same may be true for the association of these two conditions with diverticular disease of the colon, usually called Saint’s triad. 36 Requests for reprints should be addressed to J. P. C. Clinique Médicale A, Centre Hospitalier Universitaire, 1 Place V. Pauchet, 80030 Amiens Cedex, France.
REFERENCES
1. Muller, C. J. B. S. Afr. med. J. 1948, 22, 376. 2. Forster, J. J., Knutson, D. L.J. Am. med. Ass. 1958, 168, 257. 3. Palmer, E. D. Am. J. med. Sci. 1962, 244, 70. 4. Kaye, M. D., Kern, F. Lancet, 1971, i, 1228. 5. Heaton, K. W. Clins. Gastroent. 1973, 2, 67. 6. Stein, G. N., Finkelstein, A. Am.J. dig. Dis. 1960, 5, 77. 7. Cohen, S., Harris, L. D. New Engl.J. Med. 1971, 284, 1053. 8. Kreel, L. Clins Gastroent. 1973, 2, 185. 9. Admirand, W. H., Small, D. M.J. clin. Invest. 1968, 47, 1043. 10. Holzbach, R. T., Marsh, M., Olszewski, M., Holan, K. ibid. 1973, 52, 1467. 11. Bell, G. D., White, J., Dowling, R. H. in Bile Acids in Human Diseases (edited by P. Back and W. Gerok); p. 155. Stuttgart, 1972. 12. Talalay, P. Methods biochem. Anal. 1960, 8, 119. 13. Bartlett, G. R.J. biol. Chem. 1959, 234, 466. 14. Van Lerenberghe, J., Cassaigne, R. Revue fr. etud. clin. biol. 1968, 13, 541. 15. Thomas, P. J., Hofmann, A. F. Gastroenterology, 1971, 65, 698. 16. Hegardt, F. G., Dam, H. Z. ErnährWiss. 1971, 10, 223. 17. Vogt-Moykopf, I., Dietz, R., Zeidler, D. Fortschr. Med. 1972, 90, 496. 18. Schmidt, H. D. Therapiewoche, 1974, 90, 496. 19. Block, M. A., Allen, H. M. Surg. Gynec. Obstet. 1971, 132, 46. 20. Friedman, G. D. Ann. intern. Med., 1968, 68, 222. 21. Friedman, G. D. J. chron. Dis. 1966, 19, 273.
VASCULAR-LABORATORY DIAGNOSIS OF CLINICALLY SUSPECTED ACUTE DEEP-VEIN THROMBOSIS D. PRESTON FLANIGAN STEVEN J. BURNHAM
JAMES J. GOODREAU JOHN J. BERGAN
JAMES S. T. YAO Blood Flow Laboratory and Division of Vascular Surgery, Department of Surgery, Northwestern University Medical School, Chicago, Illinois 60611, U.S.A.
Doppler ultrasound, impedance plethysmography, and contrast venography were performed in 207 lower limbs suspected of harbouring deep-venous thrombosis, to clarify the diagnos-
Summary
tic value and limitations of the non-invasive methods. Doppler ultrasound and impedance plethysmography were accurate in 96% and 95% of normal limbs, respectively. In limbs with venographic evidence of thrombosis requiring treatment, Doppler ultrasound and impedance plethysmography correctly detected thrombosis in 60% and 97%, respectively. Doppler ultrasound was 97% accurate in recognising chronic venous insufficiency. Impedance plethysmography was incorrectly positive in 74% of limbs with chronic venous insufficiency which had no venographically detected thrombosis. These findings suggest that, for the accurate diagnosis of clinically suspected deep-vein thrombosis, venography is necessary only in patients with chronic venous insufficiency who have normal Doppler ultrasound tests and abnormal impedance plethysmograms and in patients with abnormal cardiac hæmodynamics. In this series, 86% of limbs would have been spared venography had non-invasive tests been used. Venography, however, remains the standard test for the detection of minor calf-vein thrombosis. A diagnostic and therapeutic schema is proposed. Introduction NON-INVASIVE tests have lately been added to the vascular surgeon’s armamentarium for the diagnosis of venous disease. In a vascular laboratory these tests can be done by technicians; the results are interpreted by a physician who prepares a report to help the primary physician decide whether diagnostic venography or
therapeutic anticoagulation are required. The blood-flow laboratory of the Northwestern versity
McGaw Medical Center
uses a
Unicombination of
22. Donaldson, R. M. Gastroenterology, 1975, 68 1608. 23. Burkitt, D. P., Painter, N. S.J. Am. med. Ass. 1974, 229, 1068. 24. Mendeloff, A.I. New Engl J. Med. 1977, 297, 811. 25. Parnis, R. O. Trans. R. Soc. trop. Med. Hyg. 1964, 58, 437. 26. Shaper, A. G., Patel, K. M. East Afr. med. J. 1964, 41, 246. 27. Biss, K., Ho, K. J., Mikkelson, B., Lewis, L., Taylor, C. B. New Engl. J. Med. 1971, 284, 694. 28. Burkitt, D. P., James, P. A. Lancet, 1973, ii, 128. 29. Bassey, O. O., Eyo, E. E., Alinhanmi, G. A. Thorax, 1977, 32, 356. 30. Dyer, N. H., Pridie, R. B. Gut, 1968, 9, 696. 31. Pomare, E. W., Heaton, K. W. Br. med. J. 1973, iv, 262. 32. Low-Beer, T. S., Pomare, E. W. ibid. 1975, i, 438. 33. Pomare, E. W., Heaton, K. W., Low-Beer, T. S., Espiner, H. J. Am. J. dig. Dis. 1976, 21, 521. 34. Painter, N. S. Proc. R. Soc. Med. 1970, 63, 144. 35. Painter, N. S., Burkitt, D. P. Br. med.J.1971, 2, 450. 36. Burkitt, D. P., Walker, A. R. P.S. Afr. med. J. 1976, 50, 2136.
332
Doppler ultrasound and impedance plethysmography for non-invasive diagnosis of deep-venous thrombosis (D.V.T.). In an attempt to improve diagnostic accuracy in venous disease, Doppler ultrasound was used to detect flow reversal in femoral and popliteal veins, a common feature of patients with chronic venous insufficiency,’1 and to detect abnormal venous hsemodynamics due to cardiac insufficiency.Both conditions can yield abnormal impedance plethysmography results.2 We have evaluated the diagnostic accuracy of the non-invasive tests in comparison with contrast venography in patients with clinically suspected acute D.V.T. Correlation between the non-invasive tests and venograms was tabulated in seven specific groups-limbs with normal venograms and those with iliofemoral thrombosis, femoropopliteal thrombosis, major calf-vein thrombosis (two or more major veins occluded), minor calf-vein thrombosis (one major vein occluded), nonocclusive venous thrombosis, and chronic venous insufficiency. The results were then analysed to decide in which group, if any, venography might be eliminated and therapeutic decisions made on the basis of non-invasive testing alone.
capacitance seen in 3 s. If abnormal results were obtained, subsequent examinations were done after repositioning of the leg to check for positional venous obstruction. Tests were done before and after a 2 min exercise period consisting of dorsiflexion and plantar-flexion of the foot. The test was considered normal if post-exercise data fell into the normal range, even if pre-exercise data were abnormal. All tests were done by a trained technician, and reports were made for subsequent interpretation by a physician without the knowledge of symptoms and signs of the patient. venous
Results Normal Doppler ultrasound examination was determined by the presence of respiratory modulations in venous velocity waveforms, normal augmented signals, and Valsalva manoeuvre. Impedance plethysmography results were based on the criteria reported by Wheeler et ap,5 (fig. 1).
Materials and Methods 207 limbs were studied by Doppler ultrasound and impedance plethysmography in patients referred to the blood-flow laboratory because of suspected D.v.T. All patients underwent standard contrast ascending venography.
Doppler Ultrasound Doppler ultrasound examination
was done by strip-chart of directional velocity patterns from the common femoral and popliteal veins with a transcutaneous directional Doppler (Parks no. 806C). Changes in venous velocity in response to Valsalva and sound augmentation manoeuvre were done as previously described. 34
recordings
Impedance Plethysmography Impedance plethysmography was performed as described by Wheeler et al. Electrode cuffs were placed on the calf and a venous compression cuff was placed on the thigh. The hip and knee were flexed 30 degrees, the hip externally rotated, and the extremity elevated 20-30 degrees on pillows. The thigh cuff 43 cm water for 45 s, thus causing a rise in capacitance. After 45 s, the cuff was rapidly deflated and the slope of the fall in venous capacitance observed, The rise in venous capacitance was then plotted against the fall in was
inflated
to
venous
Fig. 1-Nomogram for impedance plethysmography. Based on the one devised by Dr H. B. Wheelers and confirmed by 8 Hull et al.
207 limbs studied, 98 were normal by contrast venography (table). 4 of these 98 limbs showed abnormal Doppler tests, giving a specificity of 96% (94/98) and a false-positive rate of 4%. In this report, "abnormal Doppler" refers to an examination showing evidence of D.V.T. An examination showing only flow reversal suggesting chronic venous insufficiency is referred to Of
EVALUATION OF NON-INVASIVE TESTING
I.P.G.
=
Impedance plethysmography. * Limbs without D.v.T.
(207 LIMBS)
333 such. 5 of the 98 normal limbs yielded abnormal impedance examinations for a specificity of 95% (93/98) and a 5% false-positive rate. All 18 limbs having iliofemoral D.V.T. on venography also had abnormal Doppler and impedance tests, thus yielding a 100% correlation in this group. Of 28 limbs with femoropopliteal thrombosis all showed abnormal impedance studies, but only 16 (57%) showed abnormal as
Doppler findings. showing major calf-vein thrombosis (thrombosis involving two or more segments), all 11 had accompanying abnormal impedance studies while only 4 of the 11 limbs (36%) showed abnormal Doppler examOf 11 venograms
,
inations. Of the 6 limbs with minor calf-vein thrombosis (thrombosis involving one segment or seen in soleal or gastrocnemius veins) demonstrated on venography, none showed abnormal Doppler tests while 1 had an abnormal impedance examination. There were 8 limbs where non-occlusive thrombosis (floating clot) was seen in. the popliteal veins on venography. Only 1 (12%) had an abnormal Doppler test and 6 (75%) had abnormal impedance examinations. 38 limbs demonstrated chronic venous insufficiency without thrombi seen on venography. 97% of these limbs had flow reversal on Doppler examination and 74% (28 limbs) demonstrated a falsely abnormal impedance examination. Discussion Various non-invasive
techniques such as phleborheography,6 strain-gauge plethysmography,7 Doppler ultrasound,3,4 and cuff impedance plethysmography2,1,1 have been used in the diagnosis of venous thrombosis. We chose Doppler ultrasound and impedance plethysmography because both methods are technically simple and can easily be done by technicians after a short training period. Although impedance plethysmography is more sensitive than Doppler ultrasound, it is less accurate in localising the site of thrombosis. Since impedance plethysmography had often been falsely positive in chronic venous insufficiency and cardiac insufficiency in our early experience, Doppler ultrasound was added in an attempt to detect these conditions. The sensitivity and specificity of the impedance plethysmography examination in this series were not significantly different from those reported by Wheeler and others.2 5 SAlthough Doppler ultrasound was excellent in the detection of iliofemoral thrombosis, the results in femoropopliteal and major calf-vein thrombosis were worse than those recorded by others. Patency of the profunda femoris vein has been thought in some instances to mask the detection of femoropopliteal thrombosis by Doppler ultrasound. Detection of major calf-vein thrombosis with Doppler ultrasound was primarily determined
by augmentation tests. Although only 36% of limbs with major calf-vein thrombosis were detected in this series by Doppler ultrasound, 100% were detected by impedance plethysmography. Barnes et al. have reported significantly better results in detecting calf-vein thrombosis with Doppler ultrasound. Our results with Doppler ultrasound of the calf are more in accord with those of Richards et al. 10 who also used the combination of impedance plethysmography and Doppler ultrasound. The discrepancy may be related to the fact that tests were
done by technicians with subsequent physician interpretation without knowledge of clinical signs and symptoms. Subjective interpretation of various diagnostic ultrasound manoeuvres by a physician performing the tests may increase the accuracy of interpretation. The high frequency of positive impedance tests in limbs with calf-vein thrombosis that we found contrasts with Wheeler’s latest study," where impedance tests were positive in only 6 of 27 limbs with calf-vein thrombosis. Impedance testing is of limited value when early thrombi fail to impede venous outflow or when there is duplication of the femoral vein.3 Our patients had clinically suspected D.V.T. of more than one day’s duration when tested in the vascular laboratory, so most of them had propagating thrombi large enough to give a positive impedance result. The false-positive rates were 3% for Doppler ultrasound and 24% for impedance plethysmography (table). If patients with chronic venous insufficiency and minor calf-vein thrombosis are excluded, impedance plethysmography yields only a 3% false-negative rate (all in the non-occlusive group), while Doppler ultrasound yields a 40% false-negative rate. Exclusion of patients with chronic venous insufficiency also changes the frequency of false-positive impedance plethysmography studies from 24% to 5% and of Doppler ultrasound from 3% to
4%. These results suggest a clinical schema for determining the need for venography and/or anticoagulation (fig.
2):
2-Clinical schema for patients with thrombosis.
Fig.
suspected acute venous
presence or absence of chronic venous insufficiency abnormal cardiac hsemodynamics should be determined by Doppler ultrasound. Patients with abnormal cardiac hasmodynamics should proceed to venography since D.v.T. in these patients cannot be diagnosed accurately by non-invasive methods. (2) When D.v.T. is suspected clinically but both Doppler ultrasound and impedance plethysmography are normal, limbs need not be tested by venography: only 3% of these limbs will have D.v.T. which should be treated. (3) If both non-invasive tests are abnormal venography is not necessary and the patient should be treated; in this way only 5% would be treated unnecessarily. (4) Limbs with abnormal Doppler ultrasound and normal impedance plethysmography were not encountered in our series.
(1) The
or
334
(5) Non-postphlebitic limbs with normal Doppler ultrasound impedance plethysmography need not be investigated by venography and should be treated; if this is done, only 5 % will be treated unnecessarily. (6) A clinical suspicion of D.V.T. superimposed on chronic venous insufficiency is more difficult to manage. When chronic venous insufficiency is detected by Doppler, an abnormal impedance examination is of no value (74% false-positive). Thus, only patients with postphlebitic limbs normal by Doppler and abnormal by impedance testing, and patients with abnormal cardiac haemodynamics, need proceed to venography for accurate diagnosis OfD.V.T. and abnormal
This plan would allow 3% of limbs with D.V.T. to go untreated and condemn 5% of limbs without D.V.T. to anticoagulant treatment and its inherent risks. Counterbalancing these drawbacks, however, are five advantages of avoiding venography,-the significant patient discomfort, radiation exposure, and the possibility of allergic reactions to contrast medium; the risk of induc-
ing phlebitis or worsening existing phlebitis by venography (as high as 20% in one series’2); non-invasive testing is about one-third the price of venography and does not require a doctor’s time other than for interpretation ; and contrast venography does not show well the
CONTROLLED CLINICAL TRIAL OF FIVE SHORT-COURSE (4-MONTH) CHEMOTHERAPY REGIMENS IN PULMONARY TUBERCULOSIS First Report of 4th EAST AFRICAN
AND
Five 4-mo regimens of chemotherapy for tuberculosis are compared. The two regimens in which rifampicin was given throughout the 4 mo were associated with bacteriological-relapse rates of 8% in the first 6 mo after stopping chemotherapy, but the three regimens in which rifampicin was given for only the first 2 mo had relapse-rates of 24-32%. There was no evidence that the addition of pyrazinamide in the second 2 mo of chemotherapy reduced the bacteriological-relapse rate. Removal of the streptomycin from the first 2 mo appeared to reduce the bactericidal and sterilising activity of the regimen, although the differences were not statistically significant. The incidence of adverse reactions was very low with all five regimens.
Summary
external and common iliac veins in 18% of cases.8 11t In
experience
laboratory approach
to guidewith patients diagnostic clinically susIn acute D.v.T. is our 179 of 207 series, possible. pected limbs (86%) could have been spared venography. 2 limbs with D.V.T. requiring treatment would have gone undeour
a
evaluation in
tected, and 5 normal limbs would have been treated if selection criteria had been used. Contrast venography can be deleted from the diagnostic evaluation of D.V.T. only when the alternatives are similar in sensitivity and specificity, cost less, and/or carry fewer risks and complications. For epidemiological or natural-history studies of acute D.v.T., non-invasive testing, which is flow-dependent, will not detect fresh thrombus or thrombi in the soleal sinuses or calf veins, or even major veins if the thrombus is not large enough to impede venous flow. Venography or 125I-fibrinogen remain the best techniques to establish the diagnosis of these conditions. our
Supported in part by the Dr Scholl Foundation University Vascular Research Fund.
and the Northwes-
tern
Requests for reprints should be addressed to J.S.T.Y., Ward Memorial Building, Medical School, 303 E, Chicago Avenue, Chicago, Illinois 60611, U.S.A.
REFERENCES
Alexander, R. H., Nippa, J. H., Folse, R. Am. Heart J. 1971, 82, 86. Wheeler, H. B., Mullick, S. C., Anderson, J. N., Pearson, D. Surgery, 1971, 70, 20. 3. Barnes, R. W., Russell, H. E., Wu, K. K., Hoak, J. D. Surg. Gynec. Obstet. 1976, 143, 425. 4. Yao, J. S. T., Gourmos, C., Hobbs, J. T. Lancet, 1972, i, 1. 5. Wheeler, H. B., O’Donnell, J. R., Anderson, F. A. Prog. cardiovasc. Dis. 1974, 17, 199. 6. Cranley, J. J., et al. Surg. Gynec. Obstet. 1975, 141, 331. 7. Barnes, R. W., Hokanson, D. E., Wu, K. K., Hoak, J. D. Surgery, 1977, 82, 1. 2.
219. Hull, R. W., et al. Circulation, 1976, 53, 696. Hull, R., et al. New Engl.J. Med. 1977, 296, 1497. Richards, K. L., et al. Archs intern. Med. 1976, 136, 1091. Benedict, K. T., Wheeler, H. B., Patwardhan, N. A. Radiology, 696. 12. Albrechtsson, U., Olsson, C. G. Lancet, 1976, i, 723.
8. 9. 10. 11.
Study
BRITISH MEDICAL RESEARCH COUNCILS
Introduction THE high level of efficacy of several 6 mo regimens of chemotherapy for pulmonary tuberculosis has been established in studies in East Africa and in many other parts of the world, relapse-rates after the end of chemotherapy being 5% or less. 1-11It was important, therefore, to investigate regimens of even shorter duration. A pilot study in France12 of two 3-mo regimens of rifampicin, isoniazid, and streptomycin had had a relatively low overall relapse-rate of 13%. Studies in East Africa5-7 had also shown that an intensive initial four-drug phase with streptomycin, isoniazid, rifampicin, and pyrazinamide given daily for 2 mo is a very effective component of short-course chemotherapy and that the pyrazinamide plays an important part, as has been confirmed in Hong Kong.8 The present study was planned to investigate whether adding pyrazinamide to the three drugs studied in France and also increasing the total duration of chemotherapy to 4 mo would improve the results. Even if not uniformly successful, this regimen might still be applicable in developing countries, where standardduration chemotherapy with regimens known to be 100% effective in antibacterial terms produces relatively low levels of success under programme conditions, 13 default from treatment being a major reason for failure. Thus, a 4 mo regimen that is . less than uniformly successful might still constitute a major therapeutic advance. Because daily administration of streptomycin in many developing countries almost invariably requires hospital admission for the duration of treatment, a regimen not including streptomycin was also investigated. This regimen, as well as being applicable to outpatients, provides direct information about the role of streptomycin in the initial intensive phase. Patients and Methods PLAN AND CONDUCT OF STUDY
Selection of 1977, 125,
Patients
-
Patients selected were Africans aged 15-65, with previously untreated extensive pulmonary tuberculosis believed to be of
