Spinal epidural abscess--from onset to rehabilitation: case study.
Abstract: Spinal epidural abscess (SEA) is a rare condition; consequently, the clinical significance of early diagnosis often is overlooked. The challenge with SEA is not its treatment, but achieving early diagnosis before neurological symptoms occur. Once neurological deficits are present, immediate treatment must be implemented to prevent deterioration that can result in paralysis. The common SEA signs and symptoms are fever, back pain, and radicular symptoms that can progress to weakness and paralysis. Magnetic resonance imaging is the imaging of choice for diagnosis. Treatments include antibiotics and surgical evacuation of the abscess if neurologic deficits are present. Nurses play a major role in caring for SEA patients by identifying those at risk, preventing complications, and collaborating with a multidisciplinary team to prepare patients for rehabilitation.
Spinal epidural abscess (SEA) is a rare condition. Due to its rarity, the clinical significance of early diagnosis often is overlooked. The overall incidence of SEA accounts for 0.2 to 1.96 patients per 10,000 hospital admissions per year (Mandell, Bennett, & Dolin, 2000). According to Strohecker and Grobovchek's study (Reihsaus, Waldbaur, & Seeling, 2000), "The problem with spinal epidural abscess is not the treatment but early diagnosis--before massive neurological symptoms occur." SEA usually presents with back pain and may be accompanied by fever. Once neurological symptoms appear, clinical deterioration can result in paralysis unless immediate surgical and medical treatments are implemented. Nurses play an important role in caring for patients with SEA, from early diagnosis to rehabilitation. Using a case study, this article reviews the need for early diagnosis of SEA. It also reviews the pathophysiology, clinical features, medical treatments, and nursing implications related to the care of SEA patients.
The spinal epidural space is the area between the spinal cord and the vertebral body. According to Martin and Yuan (1996), the spinal epidural space is a metameric segmented structure, not a uniform space (Fig 1). Some areas of the epidural space are filled with fat and veins; in other areas, the dura is in direct contact with the bone and ligament. The posterior or dorsal and lateral part of the epidural space is larger when compared with the anterior part of the epidural space, particularly in the thoracic and lumbar areas. This difference is significant because most SEAs are located posteriorly (Martin & Yuan, 1996). In a retrospective survey by Akalan and Ozgen (2000), out of 36 patients with SEA, 20 (55%) presented with posterior SEA. Also, the most frequent location of SEA is the thoracic epidural space. In a metaa-nalysis by Reihsaus et al. (2000), of the 738 patients with SEA, 35% presented with thoracic epidural abscess, 30% had lumbar or lumbosacral abscesses, and 19% had cervical epidural abscess. This finding may be attributed to a wider epidural space in the thoracolumbar segment of the vertebral column and to the well-developed extradural venous plexus, which makes this spine segment vulnerable to hematogenous spread of infection.
[FIGURE 1 OMITTED]
The pathology of SEA is not completely understood. It seems the neurologic deficits caused by SEA are caused by mechanical compression of the nerve roots and spinal cord; however, in the Reihsaus meta-analysis (2000), several authors favored the idea of vascular inflammation due to SEA with secondary hypoxia as the main pathogenic factor because neurologic dysfunction usually is disproportionate to the extent of observed mechanical compression. A combination of both compressive and ischemic effects could be the major cause of neurologic deficits associated with SEA (Chao & Nanda, 2002).
SEA can spread via a hematogenous route from a distant infectious focus such as the skin or respiratory or urinary tracts. It also can spread via direct extension from a local process, or occur as a complication of an invasive spinal procedure. It can be caused by penetrating injuries or by extension from infected pressure ulcers or paraspinal muscle abscess (Mandell et al., 2000).
The most common risk factor is immunodeficiency, which can be caused by diabetes, alcoholism, chronic renal failure, AIDS, and malignancy. Other risk factors include infectious processes adjacent to the vertebral canal, spinal trauma, and intravenous drug abuse (Chao & Nanda, 2002).
The most common organism implicated in SEA is Staphylococcus aureus (60% to 90%), while aerobic and anaerobic Streptococcus is the second-most-common organism (18%; Mandell et al., 2000). Gram-negative organisms, such as Escherichia coli and Pseudomonas aeruginosa, account for 13% of SEA, and multiple organisms account for 10% of the cases.
The classic features of SEA include spinal and radicular pain that progress to weakness over hours to days. Back pain or neck pain with cervical spine involvement, fever, spine tenderness, progressive myelopathy with bowel and bladder dysfunction, and weakness progressing rapidly to para- and quadriplegia may occur (Zaidat & Lemer, 2002). It is important to remember that not all patients have fever on presentation, especially if they are immunosuppressed.
SEA should be considered in the differential diagnosis in patients exhibiting the classic triad of spine pain, fever, and neurologic deficits. Unfortunately, these symptoms only are present in a minority of patients.
Radiological studies are the primary means of diagnosing SEA. Enhanced magnetic resonance imaging (MRI) is the imaging study of choice for SEA diagnosis (Greenberg, 2001). The MRI should be done as soon as SEA is suspected. It rapidly and accurately identifies inflammatory loci, the extent of epidural and paraspinal involvement, the predominant location of the abscess, and, frequently, determines the surgical approach. It also is used to monitor patients undergoing treatment. The typical MRI image for an SEA patient reveals an epidural mass that may be iso- or hypointense on T1-weighted imaging (Fig 2), and hyperintense on T2-weighted imaging (Fig 3; Rigamonti et al., 1999).
[FIGURES 2-3 OMITTED]
A computed tomography (CT) myelogram typically shows compression of the thecal sac with or without a block (Rigamonti et al., 1999). ACT myelogram should be used only if MRI is unavailable or contraindicated because it carries the risk of transmitting infection to the subarachnoid space with subsequent meningitis (Joshi, Hatfield, Martin, & Taylor, 2003).
Plain films usually are normal unless vertebral osteomyelitis is present. It typically takes at least 4 to 6 weeks after onset of infection for abnormalities to show on plain films. Radiographic images may show a lyric lesion, demineralization, and disc space narrowing, which all indicate chronic infection (Akalan & Ozgen, 2000).
Laboratory studies include a complete blood count, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) levels, and blood cultures (Greenberg, 2001). Leukocytosis, or a white blood count (WBC) higher than 16.7 cells/ul, commonly is observed among patients with acute SEA. Patients with chronic SEA usually present with a normal WBC. The ESR and CRP are elevated in most patients with SEA. Blood cultures--although often negative--may help to identify the causative organisms. Lumbar puncture largely has been abandoned since the introduction of MRI because it carries a risk of spreading infection to the intrathecal compartment (Martin and Yuan, 1996).
Antibiotic therapy should be initiated as soon as the diagnosis of SEA is confirmed. Treating patients with antibiotics alone is a major advance in managing SEA. Rigamonti et al. (1999) recommended nonsurgical or medical treatment only for those with entire spinal involvement; lumbo-sacral SEA with normal neurologic exam; fixed neurologic deficit; or complete paralysis for longer than 72 hours. Nonsurgical treatment also was recommended for those who are poor surgical candidates because of an operative risk factor such as severe concomitant medical problems. Bacterial cultures obtained from blood, percutaneous aspiration of abscess, or both should be taken before antibiotic initiation. Patients on medical therapy alone must be closely monitored because neurological deterioration can occur quickly.
In a study by Sorensen (2003), 10 patients were treated only with intravenous (IV) antibiotics. Eight patients did not have any serious neurological deficits on presentation, but the remaining two were critically ill, one with tetraplegia and the other with paraplegia. Blood cultures or cultures from epidural catheters identified the causative organism for all 10 patients. All were treated with at least 3 weeks of 1V antibiotics, and continued to take oral antibiotics for as long as the CRP value was elevated. Two patients died of systemic illness despite receiving IV antibiotics, while the remaining eight had good outcomes. The author concluded that medical treatment is appropriate for patients without neurological deficit on presentation.
Nonsurgical treatment, which consists of IV antibiotics followed by serial MRI and inflammatory markers, is a debated topic. According to Harrington, Millner, and Veale (2001), medical management is appropriate in certain cases, but surgical management should be considered if a patient persistently presents with fever; WBC, ESR, or CRP remain elevated after several weeks of antibiotic therapy; MRI imaging reveals more than 50% of the thecal sac is compressed; the causative pathogen is not identified; or if persistent severe pain persists. Spinal instability, deteriorating neurological status, lack of availability of emergency spinal surgery or MRI facilities, an abscess that fails to resolve after at least 6 weeks of IV antibiotic treatment, and depressed host immune response also may necessitate surgical treatment.
Immediate spinal cord decompression, drainage of the abscess, and long-term antibiotics are the mainstays of treatment for SEA. Surgical approach and techniques are tailored to the patient, and depend upon abscess location and surrounding inflammatory tissue, as well as instability of the spinal column (Rigamonti et al., 1999). SEA surgical management goals are to preserve neurological function, prevent sepsis, permanently eradicate infection, and stabilize the spine if needed (Dietze, Fessler, & Jacob, 1997). Cultures from the surgical site typically are obtained at the time of decompression to identify appropriate postoperative antibiotic therapy.
The risks and benefits of using bone grafts and instrumentation to treat spinal infection are not clearly understood. In a retrospective study conducted by Dietze et al. (1997), 20 patients underwent vertebral arthrodesis, receiving both autologous graft and allograft for SEA. Spinal instrumentation was used on 15 patients. Parenteral antibiotics were given for 6 weeks, followed by a 3month course of oral antibiotics. There was no recurrence of spinal infection at the end of the 37-month follow-up period. The authors concluded that primary arthrodesis and instrumentation could be useful for spinal infection. They also recommended aggressive debridement of the abscess, followed by a long course of antibiotics.
Initial broad-spectrum antibiotics should treat Staphylococcus aureus, the most common organism to cause SEA. Common antibiotics used include vancomycin or nafcillin. If the patient has history of severe penicillin allergy third-generation cephalosporin or clindamycin may be used to treat gram-negative bacteria. If anaerobes are suspected, metronidazole is initiated. Once final culture reports are available, antimicrobial therapy should be optimized and curtailed, accordingly. The infectious disease service should be consulted to help plan appropriate antibiotic type and duration for individual patients.
Antibiotic therapy duration usually ranges from 6 to 8 weeks, but can last longer depending upon the patient's response to treatment. Response is based upon serial MRI, inflammatory markers, such as CRP, and neurological condition.
A peripherally inserted central catheter (PICC) is placed to prepare for long-term home antibiotic therapy. Peak and trough level of antibiotics such as vancomycin should be assessed at least once a week to help ensure an adequate therapeutic dose. Depending upon the antibiotic and other medications the patient takes, appropriate laboratory tests that check renal function or platelet counts may be necessary.
Overall SEA mortality during the past two decades has been reported at 13% to 15% (Reihsaus et al., 2000). Factors that affect outcomes and prognosis are age, level of neurological impairment on admission and before diagnosis, and presence of comorbities.
Khanna, Malik, Rock, and Rosenblum (1996) and Rigmonti et al. (1999) suggested that patients who are older than 60 years of age, those with more than 50% degree of thecal sac compression, those who have comorbid conditions (sepsis and altered immunological status because of systemic disease), and those with more than 72 hours of cord compression symptoms (bowel and bladder dysfunction, paresis, or plegia) have a poor outcome. Factors such as increased inflammatory markers, which include WBCs of more than 14,000 cells/ul on admission, continuously higher CRP level during the postoperative period, and cervicothoracic involvement also are associated with poor outcome (Soehle & Wallenfang, 2002).
Patients who developed SEAs after undergoing spinal procedures have a poor outcome as well (Reihsaus et al., 2000), possibly related to delayed diagnosis because patients may attribute their spinal or back pain to the spinal procedure rather than as a sign of spinal infection. Patients who do not have neurological deficits upon diagnosis and whose microbiological pathogen is identified through blood cultures, percutaneous aspiration, or culture of epidural catheters have a good outcome when treated medically with appropriate antibiotics alone (Sorensen, 2003).
There are no studies that compare outcomes of medical vs. surgical treatment. It would be impossible to undertake such a study because it would be unethical to withhold surgery from SEA patients who would have a good chance for recovery if surgically treated.
It takes a multidisciplinary approach to treat patients with SEA. From admission to rehabilitation, nurses play a major role in improving outcomes. They are responsible for reporting any change in physical condition from baseline that may prolong inpatient hospitalization. Nurses must properly coordinate with every discipline to ease the patient's transitions from acute care to a rehabilitation setting, and then, eventually, to their community. There are several major areas of care that require special focus, including pain, impaired physical mobility, urinary retention, nutritional deficits, prevention of deep venous thrombosis, and patient and family education regarding the disease.
Pain associated with SEA is caused by spinal cord and nerve root compression, muscle spasm, and the surgery itself. Nurses should monitor pain at least every 4 hours, continually assessing its intensity and identifying aggravating factors. Nurses should proactively administer pain medications as needed, before pain accelerates. This approach helps to effectively manage pain and evaluate the adequacy of pain modalities. Nurses also can perform important physical comfort measures, which include repositioning patients every 1 to 2 hours and maintaining proper body alignment.
Impaired Physical Mobility
Physical immobility after decompression of SEA usually is related to muscle weakness, paralysis, spasticity, or pain. Physical and occupational therapists, who are important members of the multidisciplinary team, should be consulted as soon as possible. Motor function should be assessed every 4 to 8 hours for strength, muscle tone, and gait and compare with baseline exam to identify changes. Patients should be encouraged to ambulate with assistance. Patients whose physical mobility is limited by pain should be medicated before any attempt at therapy, but overmedication must be avoided to ensure patients are not too sleepy during physical therapy.
Depending upon the duration and degree of spinal cord compression, SEA patients are at risk for urinary retention. Bladder reflexes and control of micturition from higher brain centers are lost (Hickey, 2003). This results in an atonic bladder that predisposes patients to develop urinary tract infections (UTI).
Indwelling urinary catheters usually are removed soon after surgery to prevent UTIs. Because atonic bladder is common, input and output must be maintained and reviewed. Postvoid residuals (PVR) should be measured by either bladder ultrasound or straight catheterization. If the PVR is higher than 200 ml, a urology referral usually follows. Nurses may need to teach self-catheterization if the PVR continues to be high. If patients are unable to self-catheterize, an indwelling catheter is left for at least 1 week, at which time patients usually are given a voiding trial.
Deep Venous Thrombosis
Any patient with spinal cord compression is at risk for deep venous thrombosis (DVT), which is related to prolonged bed rest, long operating times, and paralyzed limbs. Overall DVT mortality is 9% among patients with spinal cord injury (SCI; Greenberg, 2001). Signs and symptoms of DVT include calf pain, leg swelling or ipsilateral leg edema, warmth, and/or superficial venous dilation. A clinical diagnosis of DVT is not sufficiently accurate, however, because the symptoms and findings often are nonspecific or will present as an unexpected pulmonary embolism (PE).
DVT prophylaxis includes passive range of motion, early ambulation, sequential compression devices (SCDs), and low-dose anoxeparin or heparin subcutaneous every 12 hours. Despite efforts to prevent DVT, it remains one of the most common and dangerous complications of acute spinal cord injury. It usually occurs during the first 2 weeks after SCI, and it may lead to pulmonary embolism, the leading cause of death among SCI patients (Aito, Pieri, D'Andrea, Marcelli, & Cominelli, 2002). This danger has led to better DVT detection methods to help prevent PE.
The use of Doppler ultrasound screening for DVT among spinal cord injury patients continues to be controversial. In a prospective study by Alto et al. (2002), doppler ultrasounds were performed on 275 SCI patients within 3 days after injury, whenever clinically requested, and 30 to 45 days after admission. Forty-eight patients were diagnosed with DVT. Only about 35% of these patients had clinical evidence of DVT. The authors concluded that routine, repeated doppler ultrasound is an effective tool in PE prevention.
The benefits of preventing nutritional deficiency for surgical patients are well documented. Patients who have poor nutritional status usually have increased complication rates, longer hospitalizations, and higher mortality (Klein & Garfin, 1996). One has to take into account that most patients with spinal infections are nutritionally depleted before surgery. Also, most patients lose their appetites or are nauseous secondary to pain medications, antibiotics, or both. Serum albumin and total lymphocyte count are the two common parameters used to assess nutritional status. A serum albumin of less than 3.5g/dL or a total lymphocyte count of less than 2,000 cells/ul is considered to represent clinical malnutrition.
Nurses play a major role in assessing nutritional status during hospitalization. They should document nutritional intake and elicit the help of appropriate providers when patients have inadequate nutrition. They should request a nutrition service consult and a several-day calorie count to help evaluate whether patients are ingesting enough calories to meet the catabolic requirements of the postoperative period. Reasons for poor appetite should be evaluated, as well. The causes typically include depression, nausea from medications, or food-related factors, such as its consistency and taste. Nutritional supplements should be offered, and families should be encouraged to bring appetizing, home-cooked meals for recovering patients.
Patients and families should always be informed early regarding what to expect after diagnosis. Educating patients and families about the disease process, the approximate length of time the patient will need to take IV or oral antibiotics, the role of the rehabilitation program, behavior modification (e.g., smoking cessation, avoiding illicit drugs), the importance of keeping follow-up and diagnostic test appointments, and long-term outcomes is essential. Education should begin once the diagnosis is confirmed. Nurses should update patients and families daily regarding plans. Patients who are educated about their medical management usually are more eager to participate in their care. It is critical to update the interdisciplinary teams on a daily basis, as well, because each team member also cares for and educates patients and helps them transition from acute to rehabilitation settings and back to their communities.
S. A. is a 47-year-old woman who presented at the emergency department (ED) with a 2-day history of increasing back pain and bilateral lower extremity (BLE) numbness and weakness. She arrived by ambulance because, for several hours, she had been unable to walk. She denied having fever, nausea, vomiting, abdominal pain, history of trauma, or urinary and bowel incontinence. Her medical history was significant; she had pyelonephritis 6 weeks earlier. She had been treated with antibiotics at a hospital for an unknown duration. Before this admission, S. A. had been in the ED twice due to low back pain (but no BLE weakness or numbness). She was discharged with pain medication, but experienced no relief. She denied any history of diabetes, HIV, cancer, or a recent spinal procedure or surgery. Her social history was positive for tobacco and previous drug abuse; she denied alcohol use.
On physical exam, she was alert and oriented x3. She was afebrile, and her vital signs were stable. Cranial nerves II to XII were intact. On motor exam, her upper extremity exam showed 5/5 strength bilaterally. Her lower extremity iliopsoas and quadriceps were 2/5 bilaterally, hamstring strength was 2/5 bilaterally, tibialis anterior was 2/5 bilaterally, gastrocnemius strength was 2/5 on the right and 3/5 on the left, and extensor hallucis longus was 0/5 on the right and 0/5 on the left. Deep tendon reflexes were 2 + in bilateral knees, 2+ in the right Achilles tendon, and 3 + in the left Achilles tendon. She had decreased sensation to pin prick in both legs. Extensor plantar response was positive bilaterally. Rectal exam revealed decrease rectal tone.
An MRI of the thoracic spine showed an enhancing lesion involving T6 and T7 on T2 weighted images compressing the spinal cord (Fig 2). Her initial laboratory results were significant, with a CRP of 47 mg/L (< 5) and an ESR of 73mm/HR (1-39). Her CBC was normal except for her WBC, which was 12.6 K/ul. Her basic metabolic profile and urinalysis were normal.
S. A. was brought immediately to the operating room for an emergency T6-7 laminectomy with transpedicular decompression without fusion. Fungal, bacterial, and acid-fast bacillus cultures from the epidural fluid were sent for analysis. She was started immediately on IV vancomycin 1 gram every 12 hours and IV piperacillin-tazobactam 3.375 mg every 8 hours.
On postoperative day 1 (POD 1), her exam showed that her iliopsoas and quadriceps were both 2/5, tibialis anterior was 2/5, and she could wiggle her toes. She mentioned severe pain at her incision site, and her pain medications were adjusted accordingly. Her incision was dry, without redness, and staples were intact. The infectious disease service recommended continuing the present antibiotic regimen of vancomycin and piperacillin-tazobactam. They also recommended a transthoracic echocardiogram to evaluate her cardiac valves for vegetation (which could be a possible source of infection), tuberculin skin testing, and an HIV test. All three tests were negative. Physical and occupational therapists also initiated therapies. A physiatrist recommended acute rehabilitation, and the case manager and social worker were informed about the plan. A PICC line was inserted during her hospital stay. Final cultures that were taken during surgery revealed methicillin-resistant coagulase negative Staphylococcus aureus. Antibiotics were changed to vancomycin and rifampin as recommended by infectious disease services.
S. A. and her family were informed about her diagnosis and therapy. She was assured that pain medications would be given and adjusted appropriately. She was on oxycodone/acetaminophen 5/325, two tablets every 4 hours around the clock, and morphine sulfate 1 to 2 mg IV every 2 to 3 hours for breakthrough pain. On POD 2, a thoracolumbar brace was fitted for spinal support. She described 8/10 pain at her incision site, which limited her participation in physical therapy. Pain service recommended a patient-controlled analgesia (PCA) pump for her. Her urinary catheter was removed, and she successfully voided after 7 hours. Abladder scan showed a PVR of 100 cc. On POD 3, her pain was improved to 5/10. Her motor exam slowly improved; her hip flexors were 3/5, and leg extension strength was 3/5. On POD 4, her PCA pump was discontinued and she began taking sustained release oxycodone, 10 mg every 12 hours. Oxycodone with acetaminophen also was continued, but on an as-needed basis for breakthrough pain. These regimens appeared to control her pain. On POD 5, S. A. was in a melancholic mood, saying she was depressed about her condition. A psychiatric consult was ordered, and the impression was recurrent depression with low risk for suicide. The psychiatrist recommended sertraline 50 mg daily.
During this time, S. A. was not eating and her abdomen was distended. She had not had a bowel movement for 4 days despite taking two laxatives, docusate sodium and senna concentrate. She was given frequent enemas and was started on a more aggressive bowel program. A nutrition consult was ordered because her albumin was 3.3 mg/dL. She was given dietary supplements between meals to meet her caloric requirements. On POD 8, she was discharged to an inpatient rehabilitation facility. Examination on discharged showed the iliopsoas and quadriceps were both 3/5, tibialis anterior was 3/5 bilaterally, and extensor hallucis longus was 2/5 bilaterally. S. A. grew more accepting of her disease. Her medications on discharge were senna, docusate sodium, famotidine, dexamethasone (which was tapered for 2 weeks), IV vancomycin, rifampin, anoxeparin 30 mg subcutaneously twice a day, oxycodone SR, oxycodone/acetaminophen, and diazepam as needed.
SEA is a potentially life-threatening disease. Despite modern diagnostic methods, mortality has not changed for the last 2 decades. Nurses should exercise a high degree of suspicion, especially for patients who present with risk factors. Outcomes depend upon early diagnosis, immediate spinal cord decompression, and appropriate antibiotics. Nurses play a major part in helping patients recover and in preventing postoperative complications.
Editor's Note: This article received an honorable mention in the JNN 2004 Case Study Writing Contest.
Aito, S., Pieri, A., D'Andrea, M., Marcelli, F., & Cominelli, E. (2002). Primary prevention of deep venous thrombosis and pulmonary embolism in acute spinal cord injured patients. Spinal Cord, 40, 300-303.
Akalan, N., & Ozgen, T. (2000). Infection as a cause of spinal cord compression: A review of 36 spinal epidural abscess cases. Acta Neurochirurgica, 142, 17-25.
Chao, D., & Nanda, A. (2002). Spinal epidural abscess: A diagnostic challenge. American Family Physician, 65, 1341-1346.
Dietze, D. D., Jr., Fessler, R. G., & Jacob, R. P. (1997). Primary reconstruction for spinal infections. Journal of Neurosurgery, 86, 981-989.
Greenberg, M.S. (2001). Handbook of neurosurgery (5th ed.). New York: Thieme.
Harrington, P., Millner, P. A., & Veale, D. (20011. Inappropriate medical management of spinal epidural abscess (Extended report). Annals of the Rheumatic Diseases, 60, 218-222.
Hickey, J.V. (2003). The clinical practice of neurological and neurosurgical nursing (5th ed.). Philadelphia: Lippincott Williams & Wilkins.
Joshi, S. M., Hatfield, R. H., Martin, J., & Taylor, W. (2003). Spinal epidural abscess: A diagnostic challenge. British .Journal of Neurosurgery, 17, 160-163.
Khanna, R. K., Malik, G. M., Rock, J. P, & Rosenblum, M. L. (1996). Spinal epidural abscess: Evaluation of factors influencing outcome. Neurosurgery, 39, 958-964.
Klein, J. D. & Garfin, S. R. (1996). Nutritional status in the patient with spinal infection. Orthopedic Clinics of North America, 27(1), 33-36.
Mandell, G. L., Bennett, J. E., & Dolin, R. (2000). Principles and practice of infectious disease (5th ed.). Philadelphia: Churchhill Livingstone, Inc.
Martin, R.J., & Yuan, H.A. (1996). Neurosurgical care of spinal epidural, subdural, and intramedullary abscesses and arachnoiditis. Orthopedic Clinics of North America, 27(1), 125-136.
Reihsaus, E., Waldbaur, H., & Seeling, W. (2000). Spinal epidural abscess: A meta-analysis of 915 patients. Neurosurgical Review, 23, 175-204.
Rigamonti, D., Liem, L., Sampath, P., Knoller, N, Namaguchi, Y., Schreibman, D. L., et al. (1999). Spinal epidural abscess: Contemporary trends in etiology, evaluation, and management. Surgical Neurology, 52, 189-186.
Soehle, M., & Wallenfang, 72 (2002). Spinal epidural abscesses: Clinical manifestations, prognostic factors, and outcomes. Neurosurgery, 51(1), 79-85.
Sorensen, P. (2003). Spinal epidural abscess: Conservative treatment for selected subgroups of patients. British Journal of Neurosurgery, 17, 513-518.
Zaidat, O. O., & Lerner, A. J. (2002). The little black book of neurology (4th ed.). St. Louis, MO: Mosby.
Questions or comments about this article may be directed to Margaret Alvarez, MSN RN APN, 1380 East Hyde Park Boulevard, #413, Chicago, IL 60615, or via e-mail at email@example.com. She is an advanced practice nurse for the University of Chicago Hospitals, Department of Neurosurgery, Chicago, IL.