Current issues of ACP Journal Club are published in Annals of Internal Medicine


Meta-analysis: Duplex ultrasonography is accurate for most peripheral arterial occlusive disease in legs

ACP J Club. 1996 Sept-Oct;125:46. doi:10.7326/ACPJC-1996-125-2-046

Source Citation

Koelemay MJ, den Hartog D, Prins MH, et al. Diagnosis of arterial disease of the lower extremities with duplex ultrasonography. Br J Surg. 1996 Mar;83:404-9.



To determine the diagnostic accuracy of arterial duplex ultrasonography for stenotic and occlusive disease in the aortoiliac, femoropopliteal, or infragenicular vascular segments.

Data sources

Studies were identified with MEDLINE (1976 to June 1994) using the terms arterial occlusive disease, arteriosclerosis, all words starting with the phrase "claudica," vascular disease, and legs. Bibliographies of relevant studies and review articles were also checked.

Study selection

English-, German-, and Dutch-language studies were selected if they included a clear definition of the study population and the scanning technique studied. Exclusion criteria were nonadult populations; studies of neoplasms, anesthesia, wounds and injuries, or varicose veins; or duplicate publication. The criteria for high-quality studies included clear definition of the study population (essential) and of the scanning technique (essential) as well as a prospective study of consecutive patients, predefined test criteria, and independent assessment of duplex ultrasonography and angiography. Level-1 studies met all criteria; level-2 studies met at least the 2 essential criteria.

Data extraction

Data were extracted on disease and baseline characteristics; scanning techniques; criteria for outcome assessment; degree of stenosis (≥ 50% stenosis, occlusion, or either); and location of the lesion (aortoiliac, femoropopliteal, or infragenicular vascular segments).

Main results

14 studies were included: 6 were level 1 and 8 were level 2. For the aortoiliac segment assessed in 5 level-2 studies, the pooled sensitivity, specificity, and positive and negative likelihood ratios (+LR and -LR) for either stenosis ≥ 50% or occlusion were 86%, 97%, {28.7, and 0.14}*, respectively. For the femoropopliteal segment assessed in 4 level-1 studies, the pooled sensitivity, specificity, +LR, and -LR for either stenosis ≥ 50% or occlusion were 80%, 98%, {40, and 0.2}*, respectively. 6 level-2 studies of the femoropopliteal segment had a similar sensitivity, specificity, +LR, and -LR (80%, 96%, {20, and 0.21}*), respectively. For the infragenicular segment, using 2 level-1 studies, the pooled sensitivity, specificity, +LR, and -LR for either stenosis ≥ 50% or occlusion were 83%, 84%, {6, and 0.18}*, respectively.


Duplex ultrasonography provides useful information for diagnosing and planning treatment for stenotic and occlusive disease in the aortoiliac and femoropopliteal vascular segments of the legs. The technique for the infragenicular vascular segment is less useful.

Source of funding: Not stated.

For article reprint: Dr. M.J. Koelemay, Department of Surgery, G4-105, Academic Medical Center, P.O. Box 22 700, 1100 DE Amsterdam, the Netherlands. FAX 31-20-566-4440.

*Numbers calculated from data in article.


This report is a well-constructed and well-done systematic review of a diagnostic test that assesses the value of duplex arterial ultrasonography in the investigation of symptomatic peripheral vascular disease and confirms the efficacy of a technique that is now routine clinical practice for many clinicians.

Only studies of adequate quality using criteria defined by the authors were included in the assessment. This is currently only a suggested procedure for meta-analysis, not one based on formal guidelines (1).

Because sensitivities and specificities were homogeneous for the aortoiliac and femoropopliteal segments, pooled summary values for each of these parameters were presented, and +LR and -LR were calculated. Tests for homogeneity, however, are relatively conservative. The summary receiver-operator characteristic (ROC) curve, which accounts for the different values of sensitivity and specificity that must necessarily occur when different cut points are selected, can be used to combine heterogeneous values (1). The observed homogeneity merely indicates that cut points with similar sensitivities and specificities were used in the reviewed studies and indirectly reflects the uniform aims of the test in clinical usage in the different studies.

The authors suggest that the principal benefit of duplex scanning is to avoid diagnostic angiography before intervention in patients with disease proximal to the calf. Lesions with > 75% rather than ≥ 50% stenosis, however, are generally accepted as suitable for angioplasty. Clinicians may also have other diagnostic strategies, for example, identification of lesions suitable for percutaneous intervention, in patients for whom open intervention is inappropriate. In these and other circumstances, different cut points with different sensitivities and specificities are required to reflect the different intentions of the test.

Charles M. Fisher, MB, BS, BSc(Med)
Royal North Shore HospitalSydney, New South Wales, Australia


1. Irwig L, Tosteston AN, Gatsonis C, et al. Guidelines for meta-analyses evaluating diagnostic tests. Ann Intern Med. 1994; 120:667-76.