A systematic review of duplex ultrasound, magnetic resonance angiography and computed tomography angiography for the diagnosis and assessment of symptomatic, lower limb peripheral arterial disease

Abstract

Objectives: To determine the diagnostic accuracy and cost-effectiveness of duplex ultrasound (DUS), magnetic resonance angiography (MRA), and computed tomography angiography (CTA), as alternatives to contrast angiography (CA), for the assessment of lower limb peripheral arterial disease (PAD).

Data sources: Ten electronic databases were searched in April 2004, with an update in May 2005. Six key journals and bibliographies of included studies were also searched and experts in the field were consulted.

Review methods: Data extraction and quality assessment were performed in duplicate. Data were analysed according to test type and diagnostic threshold. For the economic analysis, a decision tree was developed and a probabilistic sensitivity analysis performed to incorporate statistical uncertainty into the cost-effectiveness analysis.

Results: A total of 113 studies met the inclusion criteria (including six economic evaluations). For the detection of stenosis greater than 50% in the whole leg, contrast-enhanced (CE) MRA (14 studies) had the highest diagnostic accuracy, with sensitivity ranging from 92 to 99.5% and specificity from 64 to 99%. Two-dimensional (2D) time-of-flight (TOF) MRA (11 studies) was less accurate, with sensitivity ranging from 79 to 94% and specificity from 74 to 92%. 2D phase-contrast (PC) MRA (one study) had a sensitivity of 98% and specificity of 74%. CTA (seven studies) also appeared slightly inferior to CE MRA, with a sensitivity ranging from 89 to 99% and specificity from 83 to 97%, but better than DUS (28 studies), which had a sensitivity ranging from 80 to 98% and specificity from 89 to 99%. There was some indication that CE MRA and DUS were more accurate for detecting stenoses/occlusions above the knee than below the knee or in the pedal artery. The four studies of patient attitudes strongly suggested that patients preferred CE MRA to CA. CA was considered the most uncomfortable test, followed by CE MRA, with CTA being the least uncomfortable. Half of the patients (from a sample who did not suffer from claustrophobia and had no metallic implants) expressed no preference between undergoing TOF MRA or DUS; most of those who did express a preference favoured TOF MRA. In the 55 studies identified for adverse events, MRA was associated with the highest reported proportion. However, the most severe adverse events were more common in patients undergoing CA; although these were rare for both tests. The economic evaluation showed DUS dominated the other alternatives when the whole leg was assessed, by presenting higher effectiveness at a lower cost per quality-adjusted life-year (QALY; i.e. 13,646 pounds per QALY). When the assessment was limited to a section of the leg, either above the knee or below the knee, 2D TOF MRA was the most cost-effective preoperative diagnostic strategy. The incremental cost per QALY for below-the-knee comparisons was equal to 37,024 pounds when 2D TOF MRA was compared with DUS. For above-the-knee comparisons, 2D TOF MRA presented the lowest cost and slightly lower effectiveness compared with CE MRA, with a cost per QALY equal to 13,442 pounds.

Conclusions: The results of the review suggest that CE MRA has a better overall diagnostic accuracy than CTA or DUS, and that CE MRA is generally preferred by patients over CA. Where available, CE MRA may be a viable alternative to CA. The only controlled trial suggested that the results of DUS were comparable to those of CA, in terms of surgical planning and outcome. This finding conflicts with the results of diagnostic accuracy studies, which reported poor estimates of accuracy for DUS in comparison with CA. There was insufficient evidence to evaluate the usefulness of CTA for the assessment of PAD, particularly newer techniques. The results of the economic modelling suggest that for PAD patients for whom the whole leg is evaluated by a preoperative diagnostic test, DUS dominates the other alternatives by presenting higher effectiveness at a lower cost per QALY. However, when the analysis of stenosis is limited to a section of the leg, either above the knee or below the knee, 2D TOF MRA appears to be the most cost-effective preoperative diagnostic strategy. Further research is needed into a number of areas including the relative clinical effectiveness of the available imaging tests, in terms of surgical planning and postoperative outcome.