The optimal threshold of intervention for arterial bypass grafts is still controversial. Most authorities would agree that impending failure of a graft is suggested by the following:

Gupta and coworkers recommended peak systolic velocity ratios greater than 3.4 and focal peak systolic velocities greater than 300 cm/s [42]. Similar values were also suggested by Mills et al. [19].

Bandyk created a graft surveillance risk stratification model to predict graft thrombosis as illustrated in Table 25.2 [31]. Patients with Category I lesions were hospitalized, anticoagulated, and promptly treated. Patients with Category II lesions were repaired electively within 2 weeks. Category III lesions were closely observed with serial duplex examinations and repaired if the lesions progressed in severity. Category IV lesions at the lowest risk were safely observed. Westerband and coworkers were able to support these criteria with a prospective study and demonstrate all grafts at risk for thrombosis [43].

An interesting finding in Bandyk’s series was that of lesion regression when a Category III (intermediate graft stenosis, PSV 150–300 cm/s, Vr < 3.5) lesion is discovered in the first 3 months after surgery; it may regress (30–35%), remain stable, or progress to a high-grade stenosis (40–50%) [31]. Given the variable biological behavior, it is critical to perform serial duplex studies at 4- to 6-week intervals for Category III abnormalities. These lesions will usually stabilize or progress within 4–6 months [10, 42].

As suggested by the previous study from our group, abnormal duplex findings do not always mandate further therapy [1, 45]. This is especially true if the abnormal finding is moderate PSV ratio elevation near the proximal anastomosis. We speculate that the hemodynamics at vessel bifurcations, which occurs at the typical end-to-side proximal anastomosis, is not strictly comparable to flow dynamics within the graft because of size discrepancies between the graft and native artery. Possibly, this turbulence and the resulting abnormalities in peak systolic velocity ratios at the proximal anastomosis are less predictive of graft thrombosis than the same abnormalities at other locations.

Clearly, duplex surveillance of infrainguinal bypass grafts is beneficial, but how long does the surveillance program need to continue? Most reviews have shown a definite benefit up to 2 years based on the fact that 70–80% of all graft abnormalities develop and require revision during this time period. Erickson and coworkers recommend that surveillance continue indefinitely for autogenous bypass grafts [46]. They reported that 18% of the initial interventions for a duplex-detected lesion occurred after the initial 24-month period. Sixty-three percent of theses defects occurred at an anastomosis. Although the incidence of vein graft stenosis developing decreases over time, atherosclerotic changes continue in native arteries. Another important finding to look for in older vein grafts is aneurysmal degeneration of the vein. Vein dilation is usually focal and can be associated with mural thrombus, which may warrant segmental graft revision. We support the concept of lifelong vein graft surveillance, which also gives the vascular surgeon the opportunity to monitor development of atherosclerosis in other vascular beds.

Duplex ultrasonography is the method of choice for the surveillance of infrainguinal bypass grafts. Every noninvasive vascular laboratory should continuously correlate its interpretations with arteriographic findings and clinical outcomes. Any focal peak systolic velocity >300 cm/s or a peak systolic velocity ratio >3.5 between two adjacent segments is generally accepted as a strong indicator for a focal stenosis that may threaten graft patency. Low peak systolic velocities throughout the graft (<45 cm/s), as well as lack of diastolic forward flow as evidenced by loss of biphasic Doppler signals throughout the graft, may also indicate inflow or outflow problems and warrant further investigation. Arteriography and appropriate endovascular or open surgical revision of failing grafts should be judiciously implemented by the vascular surgeon to improve long-term patency and limb salvage rates.