The United States is experiencing considerable growth in its elderly population. In 2050, the population of elderly people aged 65 years and over is projected to be 83.7 million, almost double the estimated population in 2012. As a result, an increasing portion of health-care resources will be used by the “baby boomer” generation. In the United States, approximately 505,000 hip replacements and 723,000 knee replacements were performed in 2014. It is expected that an increasing number of both elective and urgent orthopedic procedures, such as joint replacements and complex fracture repairs, will be required in older patients with multiple comorbid conditions. Therefore the number of orthopedic patients who may require postoperative critical care management is expected to increase. The critical care provider must be familiar with issues unique to these patients.
Several factors have been shown to impact morbidity and mortality in orthopedic patients. For hip replacement surgery, age has been established to be an independent predictor of mortality where patients older than 80 years of age have a mortality rate 38% higher than those less than 59 years of age. Patients with an American Society of Anesthesiologists (ASA) ≥ 3, preexisting pulmonary or cardiac conditions, renal insufficiency, or diabetes mellitus also have higher morbidity. Finally, patients undergoing spine operations have the potential for unique postoperative complications, which will be discussed separately.
Although many studies have documented the utility and benefits of joint replacement and spine surgery, there continues to be a paucity of large randomized clinical trial data on interventions to minimize postoperative complications after these procedures. The recommendations in this chapter are based mostly on small case series, meta-analyses, and retrospective reviews. This chapter will focus on the care of patients undergoing total joint or spine surgery and will not discuss care of the multiply injured trauma patient.
Care After Total Joint Arthroplasty
Several issues are unique to patients after total joint arthroplasty. These include venous thromboembolic disease, fat embolism, and complications related to the intraoperative use of cement.
Venous Thromboembolic Disease
Deep venous thrombosis (DVT) and pulmonary embolism (PE) are generally considered to be a single clinical entity referred to as venous thromboembolic disease (VTED). It is estimated that 80% to 90% of pulmonary emboli originate as lower extremity, pelvic, or caval DVTs. Certain orthopedic procedures such as total joint arthroplasty are associated with a very high rate of DVT. The reasons for this are most likely related to the extensive soft tissue dissection, inflammatory reaction in proximity to major blood vessels, and prolonged extremity immobility required after such procedures. The incidence of asymptomatic, venographically evident DVT in postoperative total joint arthroplasty patients is approximately 50% to 80% without chemoprophylaxis, but the majority of these thrombi resolve spontaneously with no clinical sequelae. Although there is minimal data with sufficient follow-up to know precisely the incidence, prevalence, or timeline for the development of these symptoms, studies have shown that patients who undergo total knee replacement have a much higher risk of symptomatic DVT (10%–15%) than those undergoing total hip replacement (3%–4%). Symptoms of DVT include edema, pain, erythema, or, less commonly, ulceration of the lower extremity. Asymptomatic DVT refers to thrombi that are evident only by screening imaging, such as conventional CT venography or duplex ultrasound. The risk of fatal PE is the same whether patients have symptomatic or asymptomatic DVTs (< 1%).
Although medical society guidelines have been established, recommendations regarding venous thromboembolism (VTE) prophylaxis can be applied uniformly to most orthopedic patients. The evidence available is low quality mainly because of the lack of adequate power, or randomized, blind, and controlled studies comparing one modality with another for the prevention of VTED. Most studies evaluating the efficacy of one VTE prophylaxis agent against another for the prevention of DVT have been nonblinded or nonrandomized. Studies evaluating various agents for the prevention of PE would be difficult to execute because of the low incidence of fatal PE. For example, a randomized prospective trial with an 80% power to demonstrate a 5% difference in efficacy of mechanical versus pharmacologic prophylaxis for prevention of PE would require 45,000 patients.
Total Hip Arthroplasty
Studies evaluating the efficacy of pharmacologic versus mechanical (e.g., foot pumps or sequential compression devices) prophylaxis against DVT after elective hip surgery have yielded conflicting findings. Small studies have shown that mechanical devices alone are efficacious in preventing asymptomatic DVT, but compliance with the use of these devices is highly variable. Other studies have shown that pharmacologic prophylaxis is superior to mechanical prophylaxis for the prevention of VTE, but these studies are low quality due to imprecision and high risk for bias. Regarding the different pharmacologic agents, low-molecular-weight heparin (LMWH) has been compared with other anticoagulants. When compared with unfractionated heparin (UFH), patients who received LMWH had fewer incidence of pulmonary embolism, DVT, and major bleeding events. Compared with patients who received a vitamin K antagonist (VKA), those who received LMWH had fewer deep vein thromboses but reported increased bleeding events. The data on factor Xa inhibitors (FXaI) such as fondaparinux, rivaroxaban, and apixaban are mixed, but overall seem to suggest a similar efficacy as LMWH for prevention of VTE and also a similar risk profile for bleeding. Evidence on dabigatran, a direct thrombin inhibitor, is also conflicting. When compared with LMWH, some randomized controlled trials showed inferior efficacy while others revealed comparable efficacy with similar bleeding rates. To date, the data on aspirin for VTE prophylaxis are mixed. Two large nonrandomized control trials compared antiplatelet agents with LMWH, whereas one study found no significant difference in PE and symptomatic DVT between the two treatments, the smaller study found a higher rate of PE in the antiplatelet group. When comparing antiplatelet agents with VKA, one randomized control trial compared the use of intermittent pneumatic compression (IPC) alone, IPC with aspirin, and IPC with VKA, and found no differences in incidence of VTED between the three groups.
Although the rates of asymptomatic DVT are similar for patients with hip fractures and those undergoing elective joint surgery, patients with hip fractures have a significantly higher rate of fatal PE (4%–10%). The reason for this difference is unknown and the exact site of the fracture (subcapital vs intertrochanteric) does not appear to affect the rate of VTED. One small randomized control trial comparing aspirin with IPC alone, found no difference in VTE events. However, pharmacologic intervention has been shown to be far superior to placebo alone. Low-dose unfractionated heparin has been shown to be efficacious in preventing VTED in small uncontrolled studies. Similarly, LMWH and warfarin have also been shown to be effective in larger studies. For hip fracture patients, fondaparinux has been shown to have higher efficacy than enoxaparin for the prevention of VTE, with conflicting results regarding the risk of bleeding complications. The Agency for Health-care Research and Quality (AHRQ) published the largest meta-analysis so far regarding prevention of VTE in major orthopedic surgeries, including hip fracture surgery. They compared different classes of chemotherapy agents.
This review concluded that the data are insufficient to assess the risks and benefits of these different interventions.
Total Knee Arthroplasty
As is the case for total hip arthroplasty, there are conflicting data regarding the various prophylactic regimens for VTED after total knee replacement. It is generally accepted that mechanical devices are less efficacious than pharmacologic agents in preventing asymptomatic DVT after knee arthroplasty. Several randomized controlled trials comparing LMWH with warfarin for the prevention of VTE after total knee replacement have failed to show a difference in PE rate, but LMWH was associated with less DVTs but slightly higher increase in bleeding events. In recent years there has been growing evidence supporting the use of FXaI after total knee arthroplasty. In a meta-analysis, Balk et al. found that FXaI were the most efficacious intervention in preventing DVT following total knee replacement, followed by the combination of LMWH and mechanical prophylaxis. As with previous studies, there were no differences between the rates of symptomatic DVT and fatal PE, most probably a result of inadequate study power to detect such a difference. Aspirin has also been compared with other chemoprophylaxis agents. When compared with FXaI, one randomized control trial showed fewer total DVTs with FXaI and no difference in symptomatic DVT when compared with aspirin. Other studies comparing aspirin with VKA and LMWH found similar results with no significant difference in total or symptomatic DVT with either group.
Recommendations for Prophylaxis Against VTED
Prophylaxis against VTED is currently based more on the danger of PE than on prevention of DVT. Previous studies from the 1970s cited the incidence of fatal PE after joint replacement surgery to be higher than 2% to 3% despite some form of prophylactic anticoagulation. Based on these papers, such operations were considered to be “high risk” for VTED and either mechanical or pharmacologic prophylaxis was encouraged. However, more recent studies have documented an incidence of fatal PE to be only 0.05% to 0.2%, regardless of whether prophylaxis of any modality is used. Possible reasons for the significant decrease in the rate of fatal PE include improved surgical and anesthetic techniques with shorter operative times, faster discharge of patients from the hospital (resulting in earlier mobility), and improved rehabilitation techniques.
Because the incidence of fatal PE is extremely low, some have suggested that pharmacologic prophylaxis is not warranted. They argue that the incidence of fatal PE is nearly equal to the incidence of complications from therapy (e.g., bleeding) and that pharmacologic intervention has not been shown to be more efficacious than mechanical measures alone in preventing symptomatic DVT or PE. Furthermore, the vast majority of deep venous thromboses resolve spontaneously, and thus the risk of bleeding associated with pharmacologic intervention may not be justified. Others, however, note that nonfatal PE and DVT carry substantial morbidity, such as heart strain, respiratory embarrassment, and thrombophlebitis, and thus the risk associated with pharmacologic prophylaxis is justified.
Patients in the intensive care unit (ICU) have limited opportunity for movement because of either the nature of their critical illness or the need for continuous monitoring. Thus it is possible that these patients will have a higher rate of fatal PE and stronger consideration should be given to providing both mechanical and pharmacologic prophylaxis against VTED.
After these guidelines were published, the use of aspirin for VTE prophylaxis in elective knee and hip arthroplasty has been looked at by many studies. One meta-analysis comparing aspirin with other chemoprophylaxis agents, either alone or in combination with mechanical prophylaxis, found a low rate of VTE with a low risk of bleeding. These results were limited by the low quality of the studies, therefore they concluded future randomized controlled trials should investigate the efficacy of aspirin. A more recent meta-analysis by Balk et al. showed that the combination of antiplatelet drug plus mechanical devices had a lower odds of DVT compared with antiplatelet drug alone, UFH heparin, and VKA.
The current guidelines, published in 2012, from the College of Chest Physicians Consensus Conference on Prevention of VTE in orthopedic surgery patients recommend pharmacologic prophylaxis to guard against VTED after total joint replacement surgery. LMWH is the agent of choice for the prophylaxis against VTED and agents such as fondaparinux, apixaban, dabigatran, rivaroxaban, UFH, adjusted-dose warfarin, or aspirin are considered second line agents ( Table 34.1 ). Similarly, The American Academy of Orthopedic Surgeons (AAOS) also recommends pharmacologic prophylaxis against VTE after hip/knee joint arthroplasty. Mechanical and chemical prophylaxis combined is recommended in patients with a past history of VTED.
|Elective total hip or knee arthroplasty |
(use at least one of the following)
|Grade 1B||Grade 1C|
|Hip fracture surgery|
a LMWH is the preferred option in both elective joint arthroplasty and hip fracture surgery. IPCD , Intermittent pneumatic compression device; LDUH , low-dose unfractionated heparin; LMWH , low-molecular-weight heparin; VKA , vitamin K antagonist.
Fat Embolism Syndrome
Fat embolism must be distinguished by the critical care provider from fat embolism syndrome (FES). Fat embolism occurs in almost all patients who have suffered multiple trauma or have had major orthopedic surgery, it is rarely symptomatic, and it is most often diagnosed incidentally on postmortem examination. On the other hand, FES has a distinct triad of signs and symptoms with a variable disease course.
Von Bergmann first described FES in 1873 while treating a patient with a femur fracture. In the orthopedic population, it occurs most often after joint replacement surgery (1%–3% incidence) and in those with multiple long bone fractures (5%–10% incidence). It is much more commonly seen in adults than in children because olein, a lipid that is more abundant in the marrow of adults, is more likely to produce emboli than palmitin and stearin, which are found mainly in younger individuals. Prolonged bedrest or delayed definitive fracture repair (greater than 48 hours) has been shown to increase the incidence of FES. The overall mortality has improved since the mid-20th century but remains at 5% to 15%, although this figure may be an underestimation of the true incidence of death resulting from FES.
FES is classically characterized by pulmonary, cerebral, and cutaneous or retinal findings. The most severe manifestations of FES are usually noted after trauma; it is unusual to have severe multiorgan dysfunction resulting from FES after joint replacement surgery. Signs and symptoms tend to occur 12 to 24 hours after injury or operation and to worsen between 72 and 96 hours after the initial insult. Nearly all patients have some degree of pulmonary impairment, tachypnea often being the first sign. Later findings are hypoxemia and increasing requirements for supplemental oxygen. Cerebral findings are present in 80% of patients and can manifest as headache, agitation, confusion, or seizure. Finally, a petechial rash localized to nondependent regions and the oral mucosa and possible retinal edema or hemorrhage occur in 40% of patients after 24 to 48 hours. The rash tends to resolve after 1 week. Other nonspecific findings may include tachycardia and fever. Mortality is most often related to severe pulmonary dysfunction, whereas morbidity is most often the result of cerebral complications.
The exact pathophysiology underlying FES remains unknown. The two leading theories are mechanical and biochemical and it is possible that the actual cause is a combination of both proposed mechanisms. The mechanical theory states that intramedullary pressure exceeds venous pressure and causes embolization of fat globules, which then lodge in pulmonary capillaries. The resulting pulmonary hypertension causes opening of the foramen ovale (if not already patent) and allows arterial embolization of other globules, which lodge in cerebral, retinal, and dermal arterioles. Because there is a 12- to 24-hour latency in the development of symptoms and there is no direct relationship between the amount of embolized fat and pulmonary distress, the biochemical theory states that the embolized fat globules initiate an inflammatory cascade in the lung leading to respiratory impairment and noncardiogenic pulmonary edema. Cerebral impairment is postulated to be caused by a similar mechanism.
Although scoring systems have been described, FES remains mostly a clinical diagnosis because no test is sensitive or specific enough to be of use. The critical care provider must have a high index of suspicion to make this diagnosis accurately. A near ubiquitous finding is hypoxemia. Many patients have PaO 2 values less than 50 mmHg on room air within 72 hours. Accordingly, arterial blood gas analysis is the only uniformly helpful test, although the results do not differentiate FES from other causes of hypoxemia. Most patients have a normal electrocardiogram. The chest radiograph is often normal initially and lags behind clinical findings, much as is noted for acute respiratory distress syndrome (ARDS). There are only isolated case reports and small series on the use of high-resolution computed tomography for the diagnosis of FES. These reports consistently show bilateral ground-glass opacities and thickening of the interlobular septa in nondependent regions of the lungs, but these findings are not specific to FES. Other tests, such as urinalysis or evaluation of cerebrospinal fluid or blood for fat cells or eosinophils, are too insensitive to be helpful. For patients with respiratory embarrassment, bronchoalveolar lavage (BAL) showing fat cells might allow FES to be differentiated from other causes, but the role of BAL has not been studied in large, prospective trials.
The treatment of FES is centered on definitive fixation of all fractures and on respiratory support to maintain adequate oxygenation. Studies evaluating therapies such as alcohol and dextran have not shown them to be of benefit. Additionally, reports on the usefulness of corticosteroids are conflicting. Corticosteroids were first used in the 1960s and their use today remains controversial, with some studies suggesting that they may be harmful. It is thought that corticosteroids may help by inhibiting the inflammatory reaction seen with FES and by decreasing the rise in plasma free fatty acids by inhibiting pulmonary lipase activity. A recent meta-analysis looked at the effect of corticosteroids in preventing FES in patients with long bone fractures. Findings suggest that use of corticosteroids may be beneficial in preventing FES and hypoxia but there was no difference in mortality. However, it is not possible to identify patients at risk of FES, thereby making prophylactic therapy not practical.
Currently, there is no consensus or recommendation on whether corticosteroids should be prescribed or on the dosage to be used. Suggested doses of methylprednisolone have ranged from 1.5 mg/kg every 8 hours for four doses, to 30 mg/kg every 2 hours for two doses, to 7.5 mg/kg every 6 hours for 12 doses.
The overall approach to the patient with FES should be based on ensuring adequate oxygen delivery to the peripheral tissues. Management involves restoring euvolemia, correcting severe anemia, monitoring indices of perfusion, and utilizing mechanical ventilation as necessary. On the basis of current knowledge, the use of corticosteroids or other medical therapies cannot be recommended.
Complications Relating to the Use of Intramedullary Cement
Traditionally, polymethylmethacrylate has been used to fix prostheses implanted during total hip replacement surgery. Most recently it is most commonly used in repair of spinal fractures. Reports of severe transient hypotension or hypoxemia related to the use of cement first appeared in 1970, but the incidence has decreased from greater than 50% to 5%. The reason for this significant decrease in incidence remains speculative because the cause of cement-related hypotension has never been determined. The decrease may be the result of improvement in anesthetic technique, such as maintaining euvolemia and oxygenation, and also the result of better prostheses, which decrease the incidence of cement or fat (bone marrow) emboli. Cement-related hypotension may be related to systemic release of polymethylmethacrylate with resulting systemic inflammation, vasodilation, and cardiac depression. Hypoxemia may result from air and fat emboli caused by forceful instillation of cement into the medullary cavity of the femur. Intraoperative transesophageal echocardiography has demonstrated numerous emboli traveling through the right heart during cement and component insertion, but relatively few patients manifest symptoms of pulmonary embolism.
It is not known what predisposes some patients to become hypotensive or hypoxemic while most remain asymptomatic. Patients at high risk of developing cement-related complications include older adults with cancer, patients with preexisting cardiopulmonary conditions, patients undergoing operations with a long-stem femoral component, and patients with hypovolemia. Unfortunately, the factors that have been shown to decrease the incidence of cement-related complications are outside the direct control of the intensivist and include maintenance of euvolemia intraoperatively, lavage of the femoral canal, and placement of a vent hole in the prosthesis.
Patients who develop complications related to the use of cement most commonly present with isolated hypoxemia that can last between 30 minutes and 48 hours. Most often, these patients have a normal chest radiograph. The need for mechanical ventilation depends on the patient’s preexisting cardiopulmonary reserve, the severity of the reaction to the cement, and other insults that may have occurred intraoperatively (such as excessive bleeding, long operative time, or inadequate resuscitation). Ries and colleagues showed a 28% increase in the pulmonary shunt fraction after instillation of cement and further demonstrated that it took up to 48 hours for the shunt fraction to return to normal. These authors recommended evaluating patients’ pulmonary function when use of cement is planned during total hip replacement surgery. Significant hypotension is noted in only 5% of patients and is usually much more transient, often lasting less than 30 minutes. All in all, cement-related complications should be considered as a diagnosis of exclusion, and physicians caring for such patients postoperatively must exclude other causes of persistent hypotension, such as myocardial infarction or hypovolemia, even in patients who are at risk of cement-related complications. Therapy is mainly supportive and centers on ensuring adequate oxygen delivery.
Care After Spine Surgery
The types of procedures performed on the spine are varied and becoming increasingly common. Lumbar fusion rates doubled from 1979 to 1990, and the highest rise was seen in those older than 60 years of age. The planned surgery may involve single or multiple levels of vertebrae using an anterior or posterior approach and may or may not require instrumentation. Some procedures require an intraperitoneal, retroperitoneal, or thoracic approach. Percutaneous and minimally invasive procedures are becoming more common. The indications for and physiologic consequences of each approach will be discussed.
Although the various operations have a low morbidity and mortality, the need for a postoperative ICU stay is increased for patients older than 60 years, when there is extensive decompression and fusion, with greater than 60 degrees in curvature of the spine (e.g., severe scoliosis), for an anterior approach to the thoracic or lumbar spine, and in the presence of other comorbidities. Comorbidities associated with the need for postoperative ICU care include preexisting myelopathy, pulmonary disease, cardiac or coronary artery disease, renal impairment, and diabetes mellitus. Preexisting myelopathy and severe scoliosis can lead to impaired pulmonary mechanics, significant pulmonary hypertension, and right ventricular failure in those older than 40 years. Furthermore, myelopathy often requires more extensive surgical repair, resulting in increased operative time, increased blood loss, and multilevel fusion.
It is not known whether an anterior or posterior spine fusion is more physiologically taxing to the patient. However, studies suggest that complications are more common and severe after a posterior approach than with an anterior approach. Nonetheless, patients may require ICU care after an anterior approach because of the volume shifts and possible respiratory dysfunction associated with violation of the thoracic or peritoneal cavity. An anterior approach may result in transiently worsening pulmonary function, whereas a posterior approach is often better tolerated from a cardiorespiratory standpoint. Complications of a posterior approach to spinal fusion include bleeding, dural tear, neurologic injury, or prolonged operative time; complications of an anterior approach include ileus, deep venous thrombosis, or respiratory impairment (in cases involving the thoracic vertebrae). The main advantage of the posterior approach is the ability to visualize the posterior spinal elements (lamina, intervertebral disks, and spinal cord) without opening a body cavity. The main disadvantage of this approach is the limited operative field provided to the surgeon, which can make instrumentation and multilevel fusion more difficult. Thus most patients requiring multilevel fusion or tumor resection undergo an anterior approach. In this circumstance, the exposure can involve retroperitoneal dissection only, or it can be much more invasive, involving a thoracoabdominal approach. A combined anterior–posterior approach is used most often when a large correction is necessary (e.g., for scoliosis) or when an anterior approach alone is insufficient to allow multilevel fusion (e.g., for osteoporosis).
Spine surgery involving multiple levels can be especially painful. Decompression of the neural elements often provides immediate pain relief to patients but the stripping of paraspinal musculature during exposure causes acute surgical pain. The implementation of a multimodal approach to pain control centered on nonnarcotic medications has led to improved outcomes across multiple categories of orthopedic patients. A multimodal approach allows improved pain control while minimizing the risk of opioid side effects, such as respiratory depression, altered mental status, constipation, ileus, and urinary retention. Some studies have demonstrated decreased fusion rates with nonsteroidal antiinflammatory (NSAID) use after spine surgery. However, level I and level II evidence has demonstrated improved pain scores and decreased opioid use without associated risk of pseudoarthrosis with short-term perioperative NSAID use in spine patients. The use of neuromodulatory agents such as gabapentin and pregabalin is also helpful. A meta-analysis of seven studies demonstrated significant decreases in pain and opioid use without major side effects. Similarly, acetaminophen should be routinely considered part of the pain regimen unless contraindicated. Continuous local anesthetic infusions postoperatively are another adjunctive therapy that can improve pain control while decreasing opioid consumption. Lastly, continuous low-dose intravenous infusion of ketamine is highly effective in minimizing postoperative narcotic requirement.
Unique Issues after Spine Surgery
Patients undergoing spine surgery, like other surgical patients, are at risk of developing congestive heart failure, arrhythmia, bleeding, and other common postoperative complications. However, spine surgery patients have a particularly increased risk for cardiopulmonary dysfunction, infection, and certain unusual complications, such as syndrome of inappropriate antidiuretic hormone (SIADH) release, injury during intubation, and recurrent laryngeal nerve injury.
Patients are at increased risk of hypoventilation after surgery involving the thoracic or lumbar spine. Surgery involving the thorax is associated with a marked decline in the 1-second forced expiratory volume (FEV 1 ), forced vital capacity, and total lung volume. Severe atelectasis is noted in 5% to 15% of patients, and 3% to 4% develop pneumonia. Thoracoabdominal procedures are associated with significant pain, which, if not treated appropriately, can lead to hypoventilation. The severe pain can last as long as 4 days, so many patients may benefit from either epidural or patient-controlled analgesia (PCA) postoperatively. Rarely, patients who have undergone a thoracoabdominal approach may have phrenic nerve injury, which may be transient (because of traction) or permanent. Unilateral phrenic nerve injury manifests radiographically as elevation of the ipsilateral hemidiaphragm and is often well tolerated in individuals with only mild to moderate preoperative pulmonary impairment. The rate of symptomatic DVT after spine surgery is 0.5% in patients receiving either mechanical or pharmacologic prophylaxis and Dearborn and colleagues found a 2% incidence of PE in such patients. PE is rarely the cause of hypoxemia in patients who have undergone spine surgery and have received VTED prophylaxis. Critical care providers must expeditiously exclude other causes of hypoxemia in these patients.
Patients with severe scoliosis are at high risk of cardiac or pulmonary failure in the perioperative period. The severe angulation of the spine results in a restrictive pattern of ventilation, with resultant hypercapnia, hypoxemia, and pulmonary hypertension, especially in those older than 40 years. Accordingly, pulmonary or right ventricular function may deteriorate acutely in the perioperative period, particularly after aggressive fluid resuscitation. These patients may benefit from cardiac performance monitoring and may require pharmacologic support to maintain adequate right ventricular function.
Many patients with scoliosis have muscular dystrophy or cerebral palsy. The majority of these patients have a cardiac abnormality of some nature. Dysrhythmia and cardiac conduction defects have also been reported in up to 50% of these patients. Such patients require telemetry postoperatively. In addition, many patients with muscular dystrophy have involvement of the bulbar muscles and are therefore at risk of aspiration postoperatively.
Acute Vision Loss
Spine patients are at risk of postoperative vision loss or disturbance resulting from prolonged prone positioning. The pathophysiology is thought to be related to inadequate vascular perfusion of essential ocular structures such as the optic nerve. Patients with ischemic neuropathy of the optic nerve typically report painless but sudden loss of vision. Ophthalmologic examination reveals swelling or pallor of the optic disc. Other causes for vision disturbance should be considered, such as corneal abrasion, stroke, and pituitary apoplexy. Patients in the prone or lateral position for an extended period of time may be at risk of orbital compartment syndrome. This ophthalmologic emergency presents with proptosis, conjunctival swelling, periocular edema, and painful loss of vision. Direct compression of the eyes causes venous congestion and increased orbital pressure, which in turn leads to decreased perfusion. Prone positioning, use of the Wilson head frame, estimated blood loss greater than 1.5 L, perioperative hypotension, and perioperative transfusion may increase the risk of perioperative vision loss. If postoperative vision loss is suspected, an ophthalmologist should be consulted immediately and hemodynamic stability should be optimized.
Intraoperative Blood Loss
Increased blood loss is expected in revision or prolonged deformity correction surgeries. Perioperative allogenic blood transfusion has been associated with surgical site infections and urinary tract infections in lumbar spine surgery, probably caused by suppression of T-cell response. As in other instances of hemorrhage, red blood cell transfusion alone can deplete clotting factors. For large-volume blood transfusions, subsequent plasma transfusion should be considered. Perioperative core temperature control is essential. Mild hypothermia can increase blood loss up to 16%. Antifibrinolytics, such as tranexamic acid and aminocaproic acid, have greatly reduced the need for allogenic blood transfusion in total joint arthroplasty. A meta-analysis of tranexamic acid (TXA) used in spine surgery in 644 patients concluded that the antifibrinolytic reduced total blood loss by a mean of 219 mL. The authors also found a decreased rate of transfusion with TXA. Some concern remains regarding the effect of these agents on the fusion mass when arthrodesis is performed.
Patients undergoing thoracic or lumbar surgery have a higher risk of infection than other surgical patients undergoing clean operations. Clean operations are nontraumatic cases that do not involve violation of hollow viscera or infected fields. Whereas discectomy with antibiotic prophylaxis is associated with a 1% risk of infection, fusion with instrumentation is associated with a 3% to 8% risk, depending on the number of levels fused and amount of instrumentation needed. It is thought that the higher risk of infection seen with increasing levels of fusion is the result of greater dissection, operative time, and blood loss. Other contributing factors include malnutrition, chronic steroid use, and preexisting distant infections (e.g., of the urinary tract).
Syndrome of Inappropriate Antidiuretic Hormone
Approximately 5% of patients undergoing spinal fusion develop SIADH postoperatively. Risk factors for developing this complication include severe blood loss, intraoperative hypotension, and rapid correction of the spine. The cause remains uncertain and therapy is directed at maintenance of normal serum sodium through water restriction. Most often, the disorder is self-limiting. Partly because of the very long operative time needed for single-stage correction, most patients with severe scoliosis are corrected in a two-stage procedure. Depending on the operation planned, the first stage can involve exposure and decompression of the cord from either an anterior or a posterior approach. The second stage, which is usually performed 5 to 7 days after the first, may involve instrumentation and can also be approached from either side. Dividing the operation into two stages might protect against the development of SIADH because blood loss and operative time for each operation are minimized. However, many studies have shown that two-stage procedures are actually associated with more net blood loss, longer net operative time, and longer hospitalization. In conclusion, the intensive care physician needs to maintain a high index of suspicion for SIADH in the patient who has undergone spine surgery and presents with hyponatremia and high urinary sodium. Patients undergoing two-stage procedures may require more ICU care than those undergoing a single-stage operation.
Ileus and Colonic Pseudo-Obstruction
Ileus occurs in 3.5% of patients undergoing lumbar spine surgery. The rate of ileus in anterior lumbar surgery is threefold higher than that of posterior lumbar surgery (7.5% vs 2.6%.). Preexisting gastroesophageal reflux disease, posterior instrumentation, and opening wedge osteotomy are risk factors for ileus in the spine patient. Other risk factors may include patients with prolonged surgeries, multiple staged surgeries, history of chronic pain, and delayed mobilization postoperatively.
It is critical that colonic pseudo-obstruction (also known as Ogilvie syndrome) is distinguished from ileus. The patient will present with obstipation in both instances, but the key feature in the former is isolated distention of the colon. In cases where a plain x-ray cannot distinguish between an ileus and colonic pseudo-obstruction, a noncontrast CT scan can be diagnostic. Although the exact cause of Ogilvie syndrome is not known, it is believed to be caused by autonomic impairment and parasympathetic inhibition. Mechanical distal colonic obstruction, such as a neoplasm, must be ruled out. Once this has been ruled out, the treatment of choice is intravenous administration of neostigmine. Colonoscopic decompression is also an option when neostigmine is contraindicated or unsuccessful. It is important that subsequent narcotics be stopped and that the patient be mobilized to prevent recurrence of either disorder.
Superior Mesenteric Artery Syndrome
Superior mesenteric artery (SMA) syndrome is a rare but known complication in patients undergoing spine surgery for deformity correction. As the alignment of the spine is corrected, the angle between the SMA and third part of the duodenum decreases. This narrowed angle may lead to external compression of the duodenum by the SMA. This complication has been most associated with the pediatric or adolescent scoliosis patient. Patients with SMA syndrome typically develop abdominal pain, emesis, and bloating 5 to 10 days postoperatively. Braun et al. identified multiple risk factors for SMA syndrome: body mass index less than 25th percentile for age, stiff thoracic curve, laterally displaced lumbar curve, thoracoplasty, anterior approach, staged procedures. The authors also emphasized the importance of nutritional optimization preoperatively. Most patients with SMA syndrome can be successfully treated with nasogastric tube decompression, repositioning the patient to a prone or left lateral position, and appropriate fluid and electrolyte replacement. Laparotomy is reserved for patients who fail medical management or who have deteriorating clinical status and concern for bowel ischemia.
In summary, the number of patients older than 65 years who require elective or urgent orthopedic surgery is expected to increase. Often, these patients have significant comorbidities that need to be addressed in the perioperative period and they may require ICU care. In addition to the routine care of the postoperative patient, those undergoing major orthopedic surgery have unique issues related to their propensity for thromboembolic disease or cardiopulmonary failure and to their need for pain management. Similarly, those undergoing spine operations are at risk of atypical postoperative complications. The astute critical care provider should be aware of these possible complications to minimize morbidity and length of stay in this patient population.