Local Anesthesia

Local anesthetics produce their effects by interfering with nerve conduction through blockade of neuronal sodium channels.^23,24 The inward sodium current is an essential component for propagation of the action potential along nerve fibers. Most local anesthetic agents in use today contain an aromatic ring and are basic and lipid soluble. The agents are made soluble in an acid aqueous vehicle for administration, but in tissue, neutral pH is required to release the basic lipid-soluble component and promote neuronal interaction and effect. This is the theoretical explanation for the diminished effect of local infiltration agents in the relatively acidic environment at sites of inflammation or infection.

Numerous specific agents are available today for local anesthesia. These agents have structural and functional homology but different physiochemical properties that influence selection for specific clinical applications, including potency, rapidity of onset, duration of action, and relative effect on sensory and motor fibers. Table 32-2 lists a number of current agents.²⁵ In clinical practice, most vascular surgeons use lidocaine or bupivacaine. Though less potent than lidocaine, bupivacaine has a longer duration of action that extends well into the postprocedural period and thus may enhance postprocedural comfort.

Local anesthetics have significant potential for toxicity.²⁶ Systemic toxicity is the result of rapid drug absorption and uptake into the circulation. Beyond the occasional inadvertent intravenous administration of these agents, systemic toxicity may occur as a result of rapid absorption from certain high–blood flow sites, such as mucosal surfaces, peritoneal or pleural spaces, and muscle. Coadministration of a vasoconstrictor (epinephrine) will reduce these effects somewhat. A Bier block, in which a relatively high dose of local anesthetic is administered intravenously into an isolated limb, can deliver a dangerous systemic dose if the tourniquet is prematurely removed. Ester-containing local anesthetics are cleared by plasma cholinesterase rapidly, but amide-based local agents, including lidocaine and bupivacaine, require liver metabolism. Toxic central nervous system effects follow a dose-related progression from vertigo and tinnitus to anxiety and fear and subsequently to tremors, myoclonic jerks, somnolence, seizures, and coma.²⁴ Benzodiazepines may mask the early neurologic signs of systemic toxicity, so particular care is needed in the setting of concomitant moderate sedation. Direct cardiovascular toxicity is seen at levels exceeding the threshold for seizure and is manifested as arrhythmia and myocardial depression.

Moderate Sedation

Moderate sedation can significantly increase patient comfort and the technical ease of invasive procedures. Moderate sedation may be administered by anesthesiology personnel in some settings but is frequently administered by proceduralists themselves with the help of specially trained and credentialed nursing staff. Such practitioners include gastroenterologists, interventional cardiologists, interventional radiologists, interventional nephrologists, pediatricians, emergency medicine physicians, and vascular specialists, among others. Clear knowledge by the practitioner of the agents used, the probable toxicities, and methods of cardiopulmonary rescue is essential for patient safety. The principal hazard of moderate sedation is induction of excessive sedation with a deeper-than-desired level of anesthesia and the potential for hypoventilation and hypoxia. The ASA has developed extensive guidelines for nonanesthesiologists who perform moderate sedation as a part of their practice.²⁰ Only a brief overview is included here.

In preparation, patients should be fasting, and a careful history of allergies, adverse reactions, current medications, and conditions recognized to compromise cardiopulmonary function under sedation (e.g., sleep apnea) should be elicited. A focused physical examination, including vital signs and auscultation of the lungs and heart, along with careful evaluation of the airway, is essential. The facility should have appropriate devices and medications for rescue in the event of oversedation and respiratory compromise. Intraprocedural and postprocedural monitoring is critical to safety. The ASA strongly recommends monitoring of the level of consciousness, oxygenation (pulse oximetry), and arterial pressure (automated oscillometry, every 5 minutes) in all cases, as well as respiration (apnea monitor) and the electrocardiogram in patients undergoing deep sedation and in those with known cardiac disease. Furthermore, it is advised that a specifically trained and designated individual other than the proceduralist be responsible for patient monitoring. Secure vascular access must be obtained and maintained through the procedure and recovery. Most physicians use a combination of sedative/hypnotic and analgesic agents administered in small, incremental doses to achieve the proper balance of effect to complete the procedure in comfort. Table 32-3 lists basic properties of several agents used in modern practice.²⁷ The pharmacology of the agents most commonly used for moderate sedation was recently reviewed.²⁸ Reversal agents (specific antagonists, such as naloxone for opiates and flumazenil for benzodiazepines) must be immediately available. These receptor antagonists may have durations of action shorter than the agents that they are used to block, and therefore continued careful observation of patients is key after any use of rescue agents.

Interest is increasing in the use of dexmedetomidine for procedural sedation. Dexmedetomidine is a highly selective, centrally acting, α₂-adrenergic agonist that induces anxiolysis, analgesia, and sedation.^28a It has a half-life of 2 to 3 hours. This agent induces sedation with limited respiratory depression and was initially approved for use as a sedative agent in adults needing mechanical ventilation for up to 24 hours.^28b Like clonidine, another centrally active α₂-agonist, dexmedetomidine has potential sympatholytic effects including hypotension, bradycardia, and decreased circulating norepinephrine, and close hemodynamic monitoring is essential with this agent. These sympatholytic effects have been shown to attenuate the stress response during emergence from general anesthesia for vascular procedures, and dexmedetomidine may be a useful adjunct in the balanced general anesthetic management of vascular patients.^28c The broad use of this agent is limited at present by its relatively high cost.

Regional Anesthesia

Regional anesthesia is usually administered by anesthesiologists, often with specific training or interest in peripheral nerves and chronic pain management. Some vascular surgeons perform regional blocks for specific procedures, such as deep and superficial cervical nerve blocks for carotid surgery and ankle blocks for minor foot procedures. The details of administration of these particular blocks can be found elsewhere in this text and in specialty texts.^29,30

Spinal/Epidural Anesthesia

Spinal and epidural anesthesia are administered by anesthesiologists, but vascular specialists must be familiar with the advantages and limitations of these approaches, particularly as they affect postoperative management and analgesia. Accordingly, a brief overview of the basic aspects is presented here. Interested readers can refer to many excellent current texts in this field.^31–33

Spinal anesthesia is contraindicated in the setting of hemodynamic instability, severe coagulopathy, increased intracranial pressure (and undiagnosed neurologic conditions that could involve increased intracranial pressure), and infection at the injection site. Deformities of the spine increase the difficulty of performing the injection but are not contraindications to this approach. The choice of agent is based primarily on the potency, rapidity of onset, and duration of action of each agent. Factors that are most important in distribution of the agent (and the eventual effect) are the dose, the position of the patient after injection, and the baricity of the solution. Anesthetics for spinal use are prepared in densities different from that of cerebrospinal fluid so that gravity can guide distribution of the agent. Lidocaine and bupivacaine are the most commonly used agents and have durations of action of approximately 60 and 100 minutes, respectively.

Spinal anesthesia produces at least some degree of sympathetic block in most patients. The major impact of this block is loss of arterial and venous tone in the affected areas along with fluid sequestration and hypovolemia. These changes with spinal anesthesia can be significant, and a decrease in systemic vascular resistance of greater than 30% has been reported in older patients with known cardiac disease.³⁴ An indwelling spinal catheter may be placed for continuous or intermittent dosing and may be a good choice for a predictably prolonged surgical procedure. Placement of the spinal catheter requires a larger-bore needle and consequently carries greater risk of post–dural puncture headache. Hypotension is generally treated by prehydration, Trendelenburg positioning, and fluid resuscitation, the sum of which can result in heart failure postoperatively if not managed carefully during the block. Vasopressors or inotropic agents may be required. A properly administered spinal block is not associated with respiratory depression in patients with normal lung function. Patients with chronic obstructive pulmonary disease may rely on the action of accessory muscles of respiration, which may be compromised by a high spinal block. Treatment of a high spinal block consists of cardiovascular and respiratory support with intubation and ventilation in some cases. Unopposed activity of parasympathetic efferents in the setting of sympathetic blockade often leads to increased gut secretions and nausea. These effects are commonly treated with atropine.

Major complications of spinal anesthesia include direct neurologic injury, cauda equina syndrome, arachnoiditis, spinal hematoma, meningitis, post–dural puncture headache, high spinal block with respiratory impairment, and idiopathic cardiovascular collapse. Fortunately, most of these complications are rare, with the exception of post–dural puncture headache, which is attributed to leakage of cerebrospinal fluid and may occur in 25% of patients after spinal anesthesia. The headache is classically relieved by the supine posture and may be associated with cranial nerve symptoms such as diplopia, tinnitus, and nausea. Patients are treated with bed rest, caffeine, hydration, and analgesics.³⁵ Symptoms may persist for 6 weeks. This complication may be treated by the administration of an epidural blood patch, which is thought to seal the meningeal puncture site.^36,37 This procedure itself is not without complications, including back pain, fever, nerve root irritation, cranial nerve palsies, and subdural hematoma.

Epidural anesthesia and analgesia in vascular patients involve placement of a catheter in the epidural space at either the lumbar or the thoracic level. Contraindications to epidural catheter placement are similar to those for spinal block, with the exception that recent anticoagulation is not an absolute contraindication. The American Society of Regional Anesthesia and Pain Medicine recently released a consensus statement suggesting that epidural blocks may be performed 4 hours after the last subcutaneous dose of unfractionated heparin and 12 hours after the last subcutaneous dose of low-molecular-weight heparin (LMWH).³⁸ Treatment with nonsteroidal anti-inflammatory drugs or aspirin (or both) is not considered a contraindication if placement is smooth, but authorities recommend discontinuance of clopidogrel 7 days before puncture. The technique of catheter placement is critical to properly enter the epidural space at the appropriate level without injury to the veins of Batson’s plexus. Local anesthetic infused into the epidural space acts on the spinal nerve roots, which are mixed nerves with somatosensory, somatomotor, and autonomic fibers. These component nerve fibers usually have different sensitivity to the blocking effect of anesthetic agents, with the greatest effect on sympathetic autonomic fibers and the least effect on somatic motor nerves. Thus the extent of autonomic, sensory, and motor blockade with a given epidural infusion may differ by several vertebral levels.

The concomitant need for preprocedural, intraprocedural, or postprocedural systemic anticoagulation may limit or complicate the use of spinal or epidural anesthetic techniques in vascular patients. With the proliferation of antithrombotic and antiplatelet agents, the guidelines for use in association with neuraxial anesthesia have become somewhat complex (Table 32-4).^38a In general, most guidelines suggest that it is safe practice to consider needle placement (or catheter removal) 4 to 6 hours after discontinuance of unfractionated heparin or 10 to 12 hours after the last dose of LMWH, and systemic heparinization 1 hour after needle or catheter insertion. Generally, elective vascular surgery should be delayed for 12 to 24 hours in the event of a traumatic tap.

The major cardiovascular, respiratory, and gastrointestinal complications of epidural anesthesia are similar to those of spinal anesthesia. Accidental dural puncture can lead to post–dural puncture headache, but this problem occurs less frequently with epidural anesthesia than with spinal techniques, which naturally always involve dural puncture. With epidural misplacement, it is possible to inject agents into either the subdural space or the subarachnoid space and potentially result in total spinal anesthesia and even block in the brainstem, which can induce substantial blockade of all cardiopulmonary reflexes, coma, and shock. This complication is extremely rare, but special care must be exercised with high epidurals.

A major advantage of catheter-based epidural anesthesia is the ability to use the catheter for postprocedural analgesia with opioids or local anesthetics, alone or in combination with either physician-controlled or patient-controlled systems. Effective epidural postoperative analgesia obviates the need for systemically administered opioids and the accompanying risk of respiratory depression, excessive sedation, and gastrointestinal side effects. Epidural anesthetic infusions can be maintained in most patients safely for 3 to 4 days.

General Anesthesia

The principles, agents, techniques, and conduct of general anesthesia are beyond the scope of this text. Comprehensive review of these aspects can be found in specialized works.³⁹ Here, only aspects of interest to vascular surgeons are considered, and coverage focuses on data-driven selection of optimal anesthetic approaches as part of the comprehensive team management of periprocedural care in complex vascular patients.

Agents used in general anesthesia, particularly inhaled anesthetics, have profound effects on the heart and circulation. Anesthetic agents induce peripheral vasodilatation and inhibit sympathetic autonomic regulation. Thus the patient has reduced ability to autoregulate the circulation and tissue perfusion and becomes dependent on therapeutic interventions (fluid and inotropes) by the anesthesia team. The loss of tonic vasoconstriction in the periphery also leads to redistribution of blood flow to the skin and loss of thermoregulation with resultant heat loss, a decrease in core temperature, and hypothermia. Similar effects are noted to a lesser degree with local sympathetic blockade induced by spinal and certain regional anesthetic techniques. The data reviewed in this text make it clear that aggressive monitoring and hemodynamic control are essential to achieve the best outcomes. Although data indicate that general anesthesia is associated with relative depression of diaphragmatic function after aortic surgery,⁴⁰ there is no clear evidence that either endotracheal intubation alone or a reasonable period of positive pressure ventilation directly induces pulmonary complications.^41,42

Aortic Surgery

The use of epidural approaches to intraoperative anesthesia and postoperative analgesia arose in the 1970s and 1980s as an attractive alternative to standard therapy with balanced general anesthesia and postoperative systemic opioid analgesia. Epidural techniques, generally supplemented with light general anesthesia during the procedure, were applied to high-risk vascular patients with apparent success. A fair number of trials were conducted in this period to compare the relative outcomes of these techniques in high-risk patients.^43–47 Unfortunately, these studies produced conflicting results, most likely because of a range of design flaws, including small sample size. Despite these collective shortcomings, a comprehensive meta-analysis combining data garnered from 141 published reports and 9559 total patients suggested a 30% reduction in mortality (odds ratio of 0.7) with the epidural technique (Fig. 32-1).⁴⁸ Similar findings were reported in a second meta-analysis by Beattie et al in 2001.⁴⁹ Since then, three major randomized clinical trials have been published that directly address this question and allow valid objective assessment of the relative merits of these different techniques. Each of these reports will be considered in some detail.

The first study is the Veterans Affairs (VA) Cooperative Study #345, reported by Park et al in 2001.⁵⁰ This prospective, randomized trial was conducted at 15 VA Medical Centers to compare specific defined outcomes in high-risk patients (ASA class III or IV) undergoing elective major abdominal surgery (gastric, biliary, colonic, and aortic). Patients were randomized to one of two anesthetic schemes: group 1 received balanced general anesthesia with isoflurane or fentanyl and postoperative pain management with systemic morphine or meperidine (intravenous or intramuscular, physician or patient controlled); group 2 received epidural anesthesia with bupivacaine combined with light general anesthesia with isoflurane or fentanyl and postoperative pain management with epidural morphine supplemented with intravenous morphine as needed. A total of 2731 patients were screened and 1021 were randomized. Most of the exclusions were for insufficient risk class, contraindication to epidural anesthesia, and unwillingness of the surgeon to randomize the anesthesia technique. The authors evaluated 10 primary and 7 secondary endpoints, including death, myocardial infarction, congestive heart failure, ventricular tachycardia, pulmonary embolism, respiratory failure, renal failure, and cerebral events. Of the 984 patients who entered the study, approximately 38% underwent major aortic surgery. The study was powered to detect a 50% reduction in relative risk of a projected 15% major complication rate. No differences were noted in the occurrence of any of the primary or secondary endpoints between the two groups. Although epidural catheters were in use for only a mean of 55 hours, pain scores were significantly less in the epidural group. A post hoc subgroup analysis of the aortic patients (n = 374) showed that occurrence of any of the 10 primary outcome events was reduced significantly in patients managed with the epidural technique (22% versus 37%), mostly as a result of a reduction in the rate of myocardial infarction, respiratory failure, and stroke. The data suggest that the two anesthetic approaches are equivalent across the range of major abdominal surgery, with some advantage seen in postoperative pain control in the epidural group. The study suffers from the lack of blinding of the treating physicians, lack of proscribed guidelines for postoperative hemodynamic management, and less-than-optimal analgesic administration in the standard-therapy group.

The second study, a single-institution study performed at Johns Hopkins Medical Center, was also reported in 2001.⁴¹ This prospective, randomized, double-blinded trial used a complete block design to assess the effects of both the intraoperative anesthesia technique and postoperative analgesic management in high-risk patients undergoing abdominal vascular surgery. Patients were randomly assigned to one of four groups: (1) general anesthesia with postoperative intravenous patient-controlled analgesia (PCA), (2) general anesthesia with postoperative epidural PCA, (3) epidural anesthesia and supplemental general anesthesia with postoperative intravenous PCA, and (4) epidural anesthesia and supplemental general anesthesia with postoperative epidural PCA. All randomized patients had an epidural catheter placed before surgery (thoracic epidural, T8-T9 or T10-T11, depending on the surgical approach), and the epidural was kept in place for at least 72 hours postoperatively. Enflurane was the inhaled agent used intraoperatively. Fentanyl was used for intravenous PCA, and combined fentanyl/bupivacaine was used for epidural infusion both intraoperatively and postoperatively. The primary endpoint was length of hospital stay, and the study was powered to detect a 20% reduction in an expected 12.7-day stay. Explicit detailed protocols were used for hemodynamic management, fluid administration, transfusion, and extubation. The study was closed prematurely after 168 patients were enrolled, when a preliminary conditional power analysis by the Data Monitoring Committee showed minimal likelihood of detecting a significant difference among the treatment groups if enrollment were to continue. The data showed no differences in the occurrence of death, myocardial infarction, myocardial ischemia, reoperation, pneumonia, and renal failure between the treatments. Similarly, ICU length of stay, time to initiation of diet, time to independent ambulation, and pain scores were not different in the four groups. Epidural PCA was associated with a significantly shorter time to extubation. These data demonstrate that in a relatively idealized, controlled setting with well-designed and well-managed treatment protocols, there are no measurable differences in critical outcomes among these techniques.

The third study to consider is the Multicentre Australian Study of Epidural Anaesthesia (the MASTER Trial), reported by Rigg et al in 2002.⁵¹ This randomized controlled trial (RCT) conducted in Australia and five other countries recruited 920 patients between 1995 and 2001. The study was intended to confirm the benefit of epidural anesthesia noted in an earlier Australian trial by Yeager et al⁴⁷ while using more consistent and modern anesthetic and analgesic technique. High-risk patients defined by a panel of specific indicators who were to undergo elective major abdominal surgery or esophagectomy were eligible for recruitment. Patients were randomized to either control, consisting of balanced general anesthesia with postoperative intravenous opioid analgesia (physician or patient controlled), or epidural procedural anesthesia with postoperative epidural analgesia (combined opioid and local anesthetic, continuous infusion). Detailed protocols were included for the management of each aspect of care throughout the study period.⁵² There was no masking of treatment in this study. The primary outcome was combined 30-day mortality and morbidity, and the study was powered to detect a 20% decrease in this parameter from an expected 50% value; 77.6% of patients were recruited in Australian centers, and 18% of patients underwent major aortic surgery (repair of abdominal aortic aneurysm [AAA]). The data showed no difference between the two treatments with regard to overall 30-day outcome (Table 32-5).^51,52 However, significant differences in pain score and respiratory failure favored the epidural group.

Taken together, these data indicate that selection of the intraoperative anesthetic technique for major abdominal vascular surgery is not a critical determinant of ultimate outcome and that equivalent overall outcomes can be obtained with either a general or an epidural technique, as recently summarized by Norris (Table 32-6).^53–57 The suggestion of benefit in the vascular subset of the VA trial⁵⁰ is counterbalanced by the carefully controlled experiment of Norris et al, which included only vascular cases.⁴¹ The epidural technique may provide some benefit in the postoperative phase, principally in terms of improved patient comfort and a reduction in pulmonary complications.^58,59 On the other hand, these studies support the greater importance of the implementation of a coordinated comprehensive care plan carried throughout the perioperative period.⁴²

Infrainguinal Reconstructions

Anesthetic technique has also been studied in patients undergoing infrainguinal vascular surgery,^60–64 but the data are less extensive and few firm conclusions can be drawn. In particular, no large-scale randomized, prospective, outcome-based studies can be used to guide anesthetic choice in this area. However, some data suggest that epidural anesthesia could have an impact on early infrainguinal bypass graft patency. Christopherson et al reported a prospective, randomized study in which general and epidural anesthesia was compared in 100 patients undergoing elective lower extremity bypass.⁶⁴ Although the data showed no evidence of an effect of anesthetic method on mortality or major morbidity, these authors found an unexpected dramatic reduction in postoperative graft thrombosis from a high rate of 43% with general anesthesia to 8% with the epidural technique. Additional investigations by the authors suggested that this result could be due to an effect of epidural anesthesia on coagulation and fibrinolysis.⁶⁵ This concept has been reviewed recently,⁶⁶ but the data are mixed. An alternative explanation for this observation may be hemodynamic, related to enhancement of graft flow secondary to reduced peripheral resistance in the setting of epidural sympathetic block. This hypothesis gains some support with data from intermittent compression after bypass⁶⁷ but remains unproven.

Carotid Surgery

The optimal mode of anesthesia for carotid surgery has been a topic of interest for decades, and numerous attempts have been made to definitively determine whether general anesthesia or local/regional techniques are superior.

Rerkasem et al⁶⁸ reviewed the published work on this question in the Cochrane Database in 2004 and concluded that there were insufficient data at that time to draw clear conclusions of benefit. The General Anesthesia versus Local Anesthesia for Carotid Endarterectomy (GALA) Trial was designed to address this data deficit, and the original report and subsequent publications have provided some useful insights. GALA was a prospective, randomized, multicenter (95 centers in 24 countries, mostly in Europe) intention-to-treat analysis comparing general anesthesia with cervical block.^68a A total of 3526 subjects (1753 subjects, general anesthesia; 1773 subjects, local anesthesia) were enrolled between June 1999 and October 2007. The primary outcome variable was composite 30-day stroke and myocardial infarction free survival. No significant difference was observed in the primary outcome between the two arms: 4.8% (84 patients) under general anesthesia and 4.5% (80 patients) under local anesthesia. Additionally, no differences were seen in the length of hospitalization or quality of life postoperatively. A subsequent publication by the GALA investigators suggested that local/regional anesthesia may be more cost-effective in patients in whom either anesthetic method is an option.^68b Thus the best anesthetic technique for the most commonly performed vascular operation remains an open question. Given the low rate of major complications with carotid surgery today, it is unlikely that a definitive answer will be forthcoming. This topic is also discussed in Chapter 100.

Capnography

Continuous monitoring of the end-tidal CO₂ using infrared absorption spectroscopy is the standard of care for the assessment of ventilation during general anesthesia. This information is displayed most commonly as a continuous plot of the partial pressure of CO₂ versus time (time capnogram). Exhaled gases can be sampled by either a sensor placed directly within the respiratory circuit (mainstream sensor, typically used in intubated patients) or a sensor placed remotely with gases aspirated via a sampling line (sidestream sensors, frequently used with nasal cannulas or face masks in patients breathing spontaneously). It is important to bear in mind that the gradient between end-tidal CO₂ and PaCO₂ (normally about 5 mm Hg) may be affected by disease states and that the measurement of end-tidal CO₂ using capnography is subject to artifacts. The clinical applications of capnography have been reviewed recently.^74a,74b In 2011, the ASA updated its statement on Standards for Basic Anesthetic Monitoring to recommend the addition of continuous monitoring of end-tidal CO₂ during both moderate and deep sedation.^74c This recommendation, although sensible, has aroused controversy because of the cost of monitoring equipment and the impact of these standards on regulators when applied to non–hospital-based treatment settings, such as dental offices and ambulatory procedural or interventional centers.^74d

Noninvasive Methods

All patients should have periodic measurement of arterial pressure with noninvasive techniques throughout the perioperative period. Today, this is generally accomplished with automated oscillometric devices that use a single cuff and measure the amplitude of oscillations as the cuff is slowly deflated. The pressure at which oscillations are maximal is the mean arterial pressure, and systolic and diastolic pressures are calculated by various proprietary algorithms.⁷⁵ Such devices correlate well with invasive arterial pressure measurements,⁷⁶ although the decreased arterial compliance in older adults and in patients with diabetes may lead to overestimation of systolic pressure and underestimation of diastolic pressure. These errors are small and of no clinical significance for intraoperative applications. Automated oscillometry may be less useful in the setting of atrial fibrillation or bradycardia and in patients with severe vasoconstriction.

Invasive Methods

Real-time measurement of arterial pressure is extremely useful in the management of anesthesia during any major procedure with the potential for rapid blood loss or other perturbations of the circulation. This is particularly the case with modern major vascular surgery, which may entail sudden changes in afterload with aortic clamping and release, intentional pharmacologic manipulation of arterial pressure (hypotension during the thoracic endograft procedure, hypertension in occasional carotid cases), and manipulation of critical cardiovascular reflex arcs (with carotid angioplasty, endarterectomy, or both). In clinical practice today, intraarterial pressure is usually measured with a fluid-filled catheter, tubing, and an electromechanical pressure transducer system. Such systems generally have sufficient dynamic response to accurately measure intraarterial pressure, at least for clinical purposes. The accuracy of these systems is limited by the length and compliance of the tubing and the frequent presence of small air bubbles in the tubing. These problems lead to a decrease in the natural frequency and an increase in damping of the system, which typically falsely increase the recorded systolic arterial pressure. More accurate measurement of intraarterial pressure can be obtained with miniaturized transducers mounted on catheters. Such devices have several drawbacks, however, including an inability to recalibrate after insertion, high cost, and poor durability, and therefore are used primarily in research applications. The currently used fluid-filled systems have the ancillary benefit of the ability to repeatedly sample arterial blood, which allows additional monitoring capability.

The arterial cannula is most commonly inserted into the radial artery. Stenosis, thrombosis, and occlusion of the radial artery are frequent complications of cannulation, and ischemia of the hand can result.⁷⁷ The ulnar artery is the dominant supply of arterial perfusion to the hand in approximately 90% of people.⁷⁸ Assessment of the adequacy of collateralization through the ulnar artery by way of the Allen test is a critical step before safe use of the radial artery for monitoring pressure. Alternatively, monitoring catheters can be placed in the brachial, axillary, femoral, or lower extremity arteries. Femoral lines can be placed at the time of arterial exposure or cannulation for diagnostic or adjunctive arteriography and can be discontinued at the end of the procedure. Brachial catheters should be avoided because of the poor collateralization around the elbow in the event of thrombosis, and axillary lines may be complicated by axillary sheath hematoma, which can produce devastating neurologic consequences if not promptly identified and treated.