ESTES guidelines: acute mesenteric ischaemia

Background: The available evidence comes from level II and III studies, with mostly small and retrospective observational case-series. However, they all consider pain as the main symptom in most cases [14, 35]. The classical presentation has been described as pain out of proportion to the findings on clinical examination [10, 19, 36‐41] but in 20 to 25 % of patients the initial presentation resembles an acute abdomen from another cause [36, 42, 43]. Other frequent symptoms associated with pain in the early course include nausea (93 %), vomiting (80 %) and diarrhoea (48 %) [19, 40, 41]. Klass’ classical description [44] of the onset of abdominal pain with sudden simultaneous passage of stools as a characteristic sign of AMI is also reported in recent studies [39, 43, 45]. However, these symptoms are not specific for AMI.

Features in the patient’s medical history can be important, particularly in the elderly presenting with 2–3 h of continuous abdominal pain [46]. A retrospective series of 215 AMI patients identified significant comorbidities that predispose to AMI such as ischaemic heart disease, atrial fibrillation, hypertension, diabetes mellitus or renal insufficiency in most patients [47]. Another retrospective series of 47 patients over 13 years observed atrial fibrillation in all 14 patients with arterial embolism and ischemic cardiomyopathy in 18 of 20 cases of arterial thrombosis [48]. As the disease progresses and ischemia leads to intestinal necrosis, pain becomes more diffuse and signs of peritoneal irritation [14] and bloating appear [39]. The patient may develop bloody diarrhoea, fever, signs of shock, multiple organ failure and a reduction in consciousness [37, 39].

Answer: Clinical assessment does not reliably distinguish between mesenteric arterial embolism or arterial thrombosis. However, the four aetiological types of AMI have been associated with different characteristics and risk factors (Table 2). EAMI is characterized by a sudden onset of pain and is frequently associated with atrial fibrillation. TAMI has a more indolent course and is often associated with a history of abdominal angina and weight loss suggestive of undiagnosed chronic mesenteric ischaemia. VAMI appears in younger patients, sometimes with several days of mild symptoms. It is associated with hypercoagulable states, cirrhosis, severe pancreatitis, abdominal trauma and advanced malignancy. NOMI is often silent as it occurs in patients that are critically ill and often ventilated.

Background: Embolism is the most frequent cause of mesenteric ischaemia (45 %) [10]. Cardiac ischemia, tachyarrhythmia, rheumatic fever, and other conditions that predispose to the formation of atrial thrombi are risk factors of the disease [35, 36, 38, 42, 49].

Approximately 33 % of patients present with a history of recent embolism [36, 39] and the absence of suitable anticoagulant treatment in these patients should increase suspicion of EAMI [21, 35, 36, 42]. The sudden onset of severe pain with spontaneous emptying of the bowel (vomiting and diarrhoea) but no significant physical findings is a classic sign of an EAMI. If a potential source of emboli can be identified, this “clinical triad” is present in 40–80 % of patients [38].

Arterial thrombosis accounts for approximately 25 % of cases of AMI [10]. These patients usually report prodromal symptoms of mesenteric angina [36] (postprandial abdominal pain, nausea and weight loss) before the acute episode resulting from pre-existent vascular insufficiency [10, 39]. The main risk factors for TAMI are atherosclerotic disease and dyslipidaemia [49, 50]. There may be a history of other vascular events and previous vascular surgery.

Venous thrombosis represents 10 % of cases of patients with acute mesenteric ischaemia and usually occurs in a younger population. While thromboembolic AMI is more common in the over 60 s, VAMI tends to occur in people over 40 [36, 47, 51, 52]. Although occasionally idiopathic, most patients have an identifiable risk factor [10, 51, 53]. Up to 50 % of patients with VAMI report a previous deep venous thrombosis or pulmonary embolism [54]. Other risk factors include hypercoagulability states (protein C and S deficiency, polycythaemia or Leiden factor V mutation), portal hypertension, abdominal trauma, abdominal infection, acute pancreatitis, malignancy, nephrotic syndrome, cirrhosis or splenomegaly [10, 40, 49, 51‐53, 55‐57]. Leiden factor V mutation is the most common associated hypercoagulability state and is reported in 20–40 % of VAMI cases [40]. Oral contraceptives, pregnancy and the puerperium are risk factors in young women [35, 58]. Venous vascular occlusion is usually peripheral, involving short segments of bowel [53]. The onset of VAMI is characterized by subacute abdominal pain that may develop over a period of up to 2 weeks. Half of the patients complain of nausea and vomiting. Untreated cases may result in portal hypertension with the development of oesophageal varices [39]. VAMI is not usually associated with postprandial syndrome, although bloating, abdominal distension, fever and occult blood in stools may be present [59].

NOMI is responsible for about 20 % of cases of mesenteric ischaemia. It usually occurs in patients that are critically ill, sedated and artificially ventilated and is difficult to recognize. It is poorly understood, but can be explained by a combination of low cardiac output and vasoconstriction. Risk factors for NOMI include age over 50, history of acute myocardial infarction, congestive heart failure, aortic insufficiency, cardiopulmonary bypass, kidney or liver disease or major abdominal surgery. Notably, many patients with NOMI may have none of these factors [10].

The diagnosis should be suspected in patients with mesenteric hypoperfusion secondary to circulatory shock or vasoactive drugs (including amines, cocaine and digitalis) when there is a significant unexpected deterioration in their clinical course [40, 41, 43, 60]. Acute or insidious pain (without defecation), bloating, abdominal distension and the presence of occult blood in the stools are all consistent with NOMI in a critically ill patient [10, 21, 36, 45, 54]. Most patients display signs of sepsis and abdominal distension as a late clinical sign [10].

Mitsuyoshi et al. [61] suggested diagnostic criteria for NOMI in the critically ill patient consisting of 3 of: ileus or abdominal pain, need for catecholamines, episode of hypotension or rising level of transaminases. Although they were able to demonstrate early detection and improved survival, the numbers involved were small.

Large doses of vasopressors alone can cause bowel ischaemia by non-occlusive, low perfusion mechanisms (NOMI). AMI has been associated with drug toxicity. Several case reports link cardiotoxins, such as digitalis, in combination with furosemide-induced fluid depletion, calcium-channel blockers [36], cocaine (linked to arterial thrombosis), organophosphates, ergotamine, phenobarbital, ethylene glycol or tricyclic antidepressants to NOMI in ICU patients. Snake bites have also been associated with NOMI [62, 63].

NOMI has also been described in patients who have undergone the stress of a surgical procedure or trauma and are receiving enteral nutrition in intensive care units. The reported incidence of AMI in these patients is 0.3–8.5 % [10, 64]. The proposed mechanism is an imbalance between demand (created by the enteral feeding) and supply (decreased by systemic hypo-perfusion and mesenteric vasoconstriction).

Background: Acute mesenteric ischaemia has been associated with several procedures and may complicate any abdominal surgery [21]. VAMI is a recognised complication of laparoscopic colorectal surgery. A retrospective analysis over 10 years (1069 colorectal operations) identified 37 (3.5 %) cases of AMI. Inflammatory bowel disease, ulcerative colitis, preoperative therapy with steroids, operative time longer than 220 min, ileoanal pouch anastomosis, total proctocolectomy or postoperative septic complications were associated with thrombosis. The latter two were independent predictors of thrombosis in multivariate analysis. Manipulation of mesenteric vessels and raised intra-abdominal pressure may play a role in the pathophysiology [65].

EAMI may occur when atheromatous plaques are dislodged by angiography of the coronary or cerebral circulation [10]. Aortic catheterization can induce cholesterol embolization [59].

AMI may be seen after colonoscopy. The decreased intravascular volume resulting from fasting and colon preparation, reduction of vascular tone through medications used for conscious sedation and the mechanical effects of colonoscopy may work together to create a low flow state precipitating acute mesenteric ischemia [66]. Possible predisposing conditions include connective tissue disease, advanced age, cardiovascular disease and immunosuppression.

AMI is an infrequent event after coronary bypass (1 %) or valve replacement surgery (0.2–0.4 %). It occurs particularly in older, dehydrated patients who have generalized atherosclerosis, and has a 70–100 % mortality [67]. Several factors (low cardiac output, use of vasopressors and underlying atherosclerotic disease) contribute to severe hypoperfusion and NOMI is the most frequent pathophysiological event [41]. Intra-aortic balloon pumps cause embolic showers particularly if placement involves excessive manipulation of a diseased aorta [50].

Background: The moribund patient with significant co-morbidities and poor performance status is unlikely to benefit from intervention. A number of factors including admission from a nursing home, or partial dependence, pre-existing ‘do not attempt resuscitation’ order, and class 4 wound before surgery are all associated with increased mortality [68]. So too are coma, artificial ventilation, acute renal failure, chronic obstructive pulmonary disease, myocardial infarction within the preceding 6 months as well as preoperative sepsis, major surgery, emergency procedure, duration of surgery and postoperative complications [68].

Most of the literature considers old age and late diagnosis as bad prognostic factors for arterio-occlusive disease. However, in VAMI longer duration of symptoms before hospital admission may be related to better outcome. With arterio-occlusive AMI a cut off point of 60-65 years and 24 h from the onset of symptoms are associated with better prognosis [7, 14, 21, 30]. One study reported thirty day survival of 81 % for patients with AMI under the age of 71. This fell to 30 % between 71 and 84 and 7 % in the over 84 s [28]. It was associated with increasing rates of non-resectable gangrene in these age groups of 9, 45 and 79 %, respectively.

In a series of patients with AMI mortality was 10.6 % if operated in the first 24 h after onset of symptoms vs. 72.9 % if operated after 24 h [14].

In arterial occlusion, all symptoms but abdominal pain indicate disease progression. The presence of peritoneal irritation is associated with bowel necrosis and worse prognosis [14, 42]. Renal failure and acidosis resulting from shock and sepsis are independent risk factors for mortality [7, 35].

In a retrospective study of 58 patients resection (at first or second look procedure) and no recent major cardiovascular intervention were associated with better survival rates [21]. A larger retrospective series of 124 patients identified older age, bandemia, elevated serum aspartate aminotransferase, increased blood urea nitrogen and metabolic acidosis as independent predictors of death in AMI [7].

APACHE-II and P-POSSUM are not accurate scoring systems for outcomes after emergency surgery and P-POSSUM may under-predict mortality from AMI [69]. However, they may provide a useful indicator of morbidity and mortality. Other scoring systems for predicting outcome from AMI have shown early promise, but require validation in further studies [2, 30, 68].

Background: The ideal plasma biomarker for AMI should be specific for bowel ischemia and highly sensitive. As ischemia starts from the mucosa and progresses to the serosa, a mucosa-derived marker would be most useful for early diagnosis [70]. The most promising plasma markers are intestinal fatty acid binding protein (I-FABP) and α-glutathione S-transferase (GST), originating from mucosa of the small bowel and D-lactate which is produced by gut bacterial organisms such as Escherichia coli [71, 72] These markers may have a potential use as early diagnostic tools in AMI, but none of them has yet entered routine clinical practice.

Serum L-lactate is a specific marker of tissue hypo-perfusion [73]. The liver can clear large quantities of L-lactate from the porto-mesenteric circulation and as a result serum lactate level does not correlate with intestinal infarction [74]. Lactic acidosis develops late in the course of AMI with extensive transmural infarction and tissue hypoperfusion due to sepsis. At that point mortality is already in the order of 75 % [75].

The fibrinolytic marker D-dimer does not differentiate patients with AMI from those with non-acute mesenteric ischemia and that there is no difference in D-dimer levels between patients with resectable and irresectable bowel necrosis [76].

The most common laboratory abnormalities found in AMI are hemoconcentration, leukocytosis, metabolic acidosis with high anion gap and lactate concentration, and high levels of serum amylase, aspartate aminotransferase, lactate dehydrogenase and creatine phosphokinase. None of these is sufficiently sensitive or specific to diagnose AMI [10, 77].

Routine laboratory tests reflect disease progression in AMI, but should not be used for diagnostic purposes (LOE: III).

A plain abdominal radiograph has no role in the early detection of AMI, since a normal radiograph does not exclude the diagnosis. Radiographic signs in patients with AMI are nonspecific and appear when bowel infarction has already occurred [78].

Percutaneous angiography has been replaced as the gold standard for diagnosis of suspected AMI by MDCT. No relevant trials supporting the use of angiography for diagnosis can be found in the recent literature [79, 80]. MDCT should be the first line imaging method in suspected AMI due to its high diagnostic accuracy [81‐85]. This imaging modality also allows other causes of acute abdominal pain to be excluded.

In this context biphasic CT involves the acquisition of scans in the arterial and venous phases. Scans are also acquired before intravenous contrast is given. Pre-contrast scans detect vascular calcification, hyper-attenuating intravascular thrombus and intramural haemorrhage, while contrast-enhanced CT allows the identification of thrombus in the mesenteric arteries and veins, abnormal enhancement of the bowel wall and the presence of embolism or infarction of other organs. Sagittal reconstructions are used to assess the origin of the mesenteric arteries [83].

The use of oral contrast is generally not feasible in patients with AMI. The transit time for oral contrast through the bowel will delay definitive treatment in AMI and the associated vomiting and an adynamic ileus limit the useful passage of oral contrast material [83, 86].

MDCT has a high specificity and sensitivity [82]. A meta-analysis of six primary studies using MDCT on 619 cases with suspected AMI showed a pooled sensitivity of 93.3 % (95 % confidence interval 82.8, 97.6 %) and a pooled specificity of 95.9 % (95 % confidence interval 91.2, 98.2 %). Among confirmed cases of AMI, final clinical diagnoses were EAMI/TAMI in 69 % of patients, VAMI in 15 % and NOMI in 16 % [87].

The high diagnostic accuracy of MDCT is based on specific features of AMI such as occlusion of mesenteric vessels and the non-specific appearance of bowel wall thickening and intestinal pneumatosis [88]. Some of these features, like diameter of the small mesenteric vessels or gas bubbles in intestinal pneumatosis, are no larger than 1–2 mm which requires a high spatial resolution achieved with MDCT with a collimation of 1.25–2.5 mm in the arterial or the portal venous phase [83, 88]. Pneumatosis Intestinalis (presence of air within the bowel wall) usually indicates transmural infarction in AMI, particularly if it is associated with portomesenteric venous gas, but this sign is not specific for either infarction or ischemia.

NOMI represents the most difficult diagnostic challenge in AMI; usually these critically ill patients are sedated and intubated and diagnosis may rely only on clinical suspicion and imaging studies. The administration of intravenous contrast may be contraindicated, hence a lesser sensitivity and utility of MDCT [89]. The diagnosis of NOMI should be confirmed by selective catheter angiography [90, 91]. On CT, the bowel wall of the involved segments may be normal or thickened and the pattern of enhancement is variable ranging from absent or diminished to increased, while fat stranding of the mesentery and ascites are usually visible [83].

The CT findings in NOMI are poorly understood [86] although in two published papers, with a total of 6 cases, that specifically addressed the CT findings of surgically and pathologically proven NOMI [92, 93] the reported CT findings consisted of diffuse narrowing of the SMA, poor visualization of its secondary branches, bowel distension, thin and poorly enhancing bowel wall, and intestinal pneumatosis which are similar to the findings of acute occlusive AMI, except for the absence of thrombus/embolus. The most common features associated with AMI are described in Table 3.

Percutaneous angiography should not be used for initial diagnosis of AMI except where NOMI is suspected (LOE: III).

Background: Several articles with small series of patients address the use of laparoscopy in AMI [94‐96]. In cases of uncertainty regarding the diagnosis or the extent of bowel necrosis, laparoscopy can offer a minimally invasive way to asses it, even in the ICU (bedside laparoscopy). Laparoscopic second-look procedures can be offered in a planned or on-demand fashion. However, there is no strong evidence to support the routine use of laparoscopy in AMI.

Background: The principal goals of managing the patient with AMI can be summarized with the 3 “Rs”: Resuscitation, Rapid diagnosis and early Revascularization before there is significant progression of the systemic inflammatory response. The main goal of fluid resuscitation is the restoration of adequate tissue/organ perfusion [97‐99]. Supplementary oxygen should be given [100] and the evaluation of organ delivery of oxygen assessed using accepted clinical markers such as peripheral perfusion, mental status, and urine output [97, 100].

A recent Cochrane Review [101] has shown no advantage of colloids over crystalloids in reducing mortality when used for intravenous fluid resuscitation. The choice of fluid will have a cost implication, and colloids are more expensive than crystalloids. As there is no evidence that colloids improve survival rates or reduce morbidity, crystalloids should be preferred on economic grounds. Evidence suggests that the use of hydroxyethyl starch might increase mortality and should be avoided until further proof of safety can be provided [101‐105].

Fluid volume status should be quickly assessed and fluid replacement should start promptly but should not delay diagnosis and intervention (LOE: IV).

Crystalloids should be used for fluid replacement, hydroxyethyl starch should be avoided (LOE: Ia).

Background: AMI results from insufficient blood flow within the mesenteric circulation to meet the metabolic demands placed upon it. This may be confounded by hypovolaemia and circulatory failure. The use of vasopressors to improve cardiac function should be balanced against the adverse effect of further splanchnic vasoconstriction [3, 106]. Consideration should be given to the use of drugs such as dobutamine, low dose dopamine and milrinone, which have been shown to have less impact on mesenteric blood flow [39]. Vasopressors should not be used until the patient has been adequately volume resuscitated. Cardiac output can also be improved by rate control in atrial fibrillation. However, digoxin and other cardiac glycosides also reduce flow in the splanchnic circulation and should be avoided for the control of atrial fibrillation/flutter in patients with AMI [40, 96, 98].

Cardiac glycosides should not be used as first line treatment of atrial fibrillation/flutter in AMI (LOE: IV).

Background: Loss of the integrity of the mucosal barrier facilitates bacterial translocation and occurs in the early stages of AMI. Although no specific studies have examined the role of prophylactic antibiotics in AMI, broad spectrum antibiotics (such as a penicillin or a third generation cephalosporin in combination with metronidazole) would be expected to reduce the consequences of bacterial translocation and should be given early [107].

Background: Across the literature there is agreement that peritonitis secondary to bowel necrosis mandates surgery without delay if curative treatment is to be achieved [98, 106]. All other priorities (type of AMI, level of occlusion, etc.) are secondary and should be managed accordingly.

Patients considered unsalvageable should have palliative care (LOE: IV).

Background: Patients with AMI who have severe sepsis or septic shock and undergo life-saving surgery should have a damage control approach. [118‐121]. This includes immediate laparotomy with resection of ischaemic bowel (and no anastomosis or stoma), open thrombectomy (if indicated), transfer to the intensive care unit to continue resuscitation prior to planned definitive procedures [122]. A temporary abdominal closure via a negative pressure wound therapy (NPWT) device should be considered since they have been shown to promote wound healing and facilitate subsequent abdominal closure [123, 124]. A scheduled ‘second look’ procedure should be performed within 48 h [125, 126].

Background: Necrotic bowel will be clearly identifiable at laparotomy [122, 127]. The viability of the remaining bowel should be determined after the patient has been adequately resuscitated and any resection/revascularization performed [122]. In the acute setting, hypotension, vascular impairment and the concurrent use of vasopressors may confound this assessment and could result in excessive resection [106]. It is often better to adopt a damage limitation approach and schedule a second-look procedure under these circumstances.

Intra-operative assessment with Doppler ultrasound of the vascular arcade, [128, 129], fluorescein angiography [127, 130] and indocyanine green angiography [131] have shown promise, but not been used extensively.

Ischemic bowel should be reassessed after adequate fluid resuscitation and revascularization. If any doubt remains about the viability of the bowel, a second-look procedure should be performed (LOE: IV).

Background: Vascular occlusion in TAMI usually occurs in the proximal SMA and the consequent ischemia may extend from the terminal portion of the duodenum to the transverse colon [36]. In contrast, in EAMI most emboli lodge in the superior mesenteric artery distal to the origin of the middle colic artery. This maintains perfusion of the inferior pancreaticoduodenal arteries and spares the proximal jejunum from ischaemia [10, 36, 132]. However, up to 15 % of emboli in EAMI lodge at the origin of the SMA causing more extensive ischaemia.

Short bowel syndrome requiring total parenteral nutrition (TPN) has been reported in 13–31 % of long-term survivors of AMI [9, 21, 31] and weight loss has been documented in 38 % [31]. SBS generally occurs when less than 200 cm of functioning bowel remains [106].

Restoring bowel continuity will improve the functional results of extensive small bowel resection and may avoid long term TPN. It has been suggested that while permanent intestinal failure is likely with a residual 100 cm of small bowel and an end jejunostomy, the need for permanent TPN may be avoided with a minimum 65 cm of jejunum and a jejunocolic anastomosis, and 35 cm where the ileocaecal region is preserved and a jejunoileal anastomosis performed [133].

When a residual small bowel length of 200 cm cannot be obtained, careful consideration should be given to the significant risk of resection, particularly in the elderly patients and in those with significant co-morbidities. In younger patients the option of long term parenteral nutrition or the option of subsequent intestinal transplantation could make a more extensive resection reasonable.

Background: Bowel anastomosis should be avoided in the presence of severe sepsis or septic shock or when the patient is inadequately resuscitated. Where limited necrosis is found, there is no doubt about the viability of the remaining bowel, the patient has been adequately resuscitated and there is no evidence of shock, some authors advocate anastomosis with a planned second-look [94]. Laparoscopy can be useful in this circumstance [126]. There are no published studies that compare anastomotic techniques in AMI.

Stomas avoid the risks of anastomotic failure and permit easy examination of the bowel by inspection and/or endoscopy [106]. A mucous fistula of the distal bowel may be used for re-feeding in patients with very proximal jejunostomies [134]. Continuity can be restored later if appropriate.

If during a second look procedure there is still doubt about the viability of the bowel it should be exteriorised (LOE: IV).

Background: The use of second look procedures is associated with a significant reduction in morbidity (p = 0.002) [9] and mortality in selected patients leading to resection of ischaemic bowel in 38–71 % of cases [14, 17, 18, 24, 29, 126, 135]. It is currently the best method to assess bowel viability after revascularization and resuscitation and may be the best time to perform an anastomosis and to close the abdomen after DCS.

A scheduled second-look should also be considered when there is concern about the possible progression of bowel ischemia and there should be a low threshold for an “on demand” second look procedure in AMI particularly when a bowel anastomosis has been performed (LOE: III).

Background: Mortality increases dramatically when symptoms have been present for more than 24 h in AMI. Numerous studies have shown that mortality is lowest if intervention is performed within 12 h of the onset of symptoms. [18, 22, 25, 58, 136‐139]. This is supported by experimental work which has shown gut viability to be 100 % within the first 12 h of ischaemia. This drops to 54 % between 12 and 24 h and 18 % beyond 24 h [140].

The development of symptoms in VAMI is usually slower than the other forms of AMI. A small retrospective study of VAMI over 23 years suggested that patients presenting within 3 days of the onset of symptoms had poorer outcomes than those presenting with symptoms of longer duration. The first group were more likely to undergo laparotomy within 12 h of hospital admission (83 vs. 20 %) [52]. This reflected rapid disease progression in the group that presented sooner. In studies comparing the four aetiological types of AMI, VAMI has the lowest mortality (11–30 %) [5, 22, 30].

Background: Since TAMI is usually a marker for advanced systemic atherosclerosis, it would seem reasonable, where appropriate, to advise these patients to stop smoking, modify their diet, lose weight and exercise adequately. Diabetes and hypertension should be treated and the use of statins and antiplatelet therapy considered [141].

Up to 60 % of patients with TAMI have previous symptoms of chronic mesenteric ischaemia (CMI) [9]. The mortality of elective surgical intervention for CMI in selected patients has been shown to be significantly lower than the same procedures performed for TAMI [32]. However, there is little evidence that elective intervention for CMI prevents TAMI.

One-third of patients with EAMI have inadequately treated atrial fibrillation at presentation [9] and a similar number will have further emboli [3]. These patients should be considered for anticoagulation [31].

Cho et al. advocate ‘vigilant surveillance and preservation of SMA patency’ after vascular reconstruction [32]. Although numbers were small, 60 % (3 of 5) of patients who were treated for graft stenosis with stent/angioplasty or further reconstruction had a successful outcome. However, all of these occlusions were symptomatic.

VAMI is usually secondary to a hypercoagulable state [3]. These patients should be investigated for thrombophilia and treated accordingly.

Patients with proven CMI should be considered for elective revascularization (LOE: IV).

Unless contraindicated, patients with emboli should be treated with life-long anticoagulation to reduce the risk of recurrence (LOE: IV).

Patients with mesenteric artery thrombosis are at high risk of coronary thrombosis and they should be offered anticoagulants or antiplatelet therapy and a statin to reduce this risk (LOE: IV).

Patients with mesenteric venous thrombosis should be investigated for thrombophilia and treated accordingly. Where indicated, they should have a minimum of 6 months of anticoagulation (LOE: III).

Patients who have vascular stents or bypass procedures should have graft surveillance or stent follow up with duplex scanning or CT angiography to enable early detection and management of stenosis and occlusion (LOE: IV).