Surg Neurol 1992 ;38 :225-31
225
Spinal Epidural Abscess : A Report of 40 Cases and Review Eric S . Nussbaum, M.D., Daniele Rigamonti, M .D., Harold Standiford, M .D., Yuji Numaguchi, M.D., Aizik L . Wolf, M.D., and Walker L . Robinson, M .D. Departments of Surgery, Division of Neurosurgery ; Diagnostic Radiology, Division of Neuroradiology ; and Internal Medicine, Division of Infectious Disease, University of Maryland Medical Systems, Baltimore Maryland
Nussbaum ES, Rigamonti D, Standiford H, Numaguchi Y, Wolf AL, Robinson WL. Spinal epidural abscess : a report of 40 cases and review . Surg Neurol 1992 ;38 :225-31 . Despite modern medical advances, the morbidity and mortality rates associated with spinal epidural abscess remain significant, and the diagnosis often is elusive . A retrospective study was undertaken to define better the incidence and clinical features of this infection, and to establish current diagnostic and therapeutic guidelines . Forty cases of spinal epidural abscess were encountered at our institution between July 1979 and March 1991 . All medical records and radiological images were reviewed . We report a significant increase in the incidence of epidural abscess after June 1988 (p = 0 .0195). Sixteen patients used drugs intravenously, and six had undergone spinal procedures. Twelve patients were misdiagnosed in various emergency rooms or clinics and discharged . Localized back pain, fever, and neurological deficit remained the typical clinical manifestations . Erythrocyte sedimentation rate was elevated uniformly when measured (21 cases) . Magnetic resonance imaging was diagnostic specifically in 23 of 24 instances . The majority of patients underwent surgical drainage, but five selected patients were managed nonoperatively . The highly variable presentation of spinal epidural abscess may confuse the diagnosis and delay indicated surgical intervention . Localized back pain in a febrile patient at significant risk for epidural abscess warrants erythrocyte sedimentation rate measurement . The presence of erythrocyte sedimentation rate elevation or evidence of spinal cord compression on physical examination are indications for immediate magnetic resonance imaging examination with contrast enhancement . Surgical drainage with sustained intravenous antibiotic treatment remains the cornerstone of therapy . Nonoperative management may be considered in selected cases . KEY WORDS : Spinal epidural abscess ; Spinal infection ; Spinal cord compression
Address reprint requests tar Daniele Rigamonti, M .D., University of Maryland Hospital, Division of Neurosurgery, Rm . NI2E27 22 South Greene Street, Baltimore, MD 21201 . Received November 18, 1991 ; accepted January 27, 1992 .
C 1992 by Elsevier Science Publishing Co ., Inc.
In 1948, Heusner studied 20 patients with spinal epidural abscess and emphasized the importance of early diagnosis [131 . Later, Baker and colleagues reviewed the clinical course and pathophysiology of epidural abscess [2] . Subsequent publications have stressed the potentially rapid deterioration and persistent high morbidity associated with this serious, localized infection [8,11,14,26,28,31,33-35] . Currently, the diagnosis of epidural abscess remains difficult, and the consequences of delayed intervention are still potentially devastating. However, since the time of the review by Baker et al, new imaging modalities have become available, more effective antimicrobial agents have been developed, and different risk factors have emerged as significant. Given these advances, our experience with spinal epidural abscess from 1979 to the present has been reviewed in an attempt to clarify issues such as incidence of epidural abscess and predisposing factors, diagnostic guidelines, treatment regimens, and outcome .
Methods and Materials From July 1, 1979, to April 1, 1991, the University of Maryland Medical Systems Hospitals admitted 245,643 patients . A careful review of patient records revealed 40 cases of documented spinal epidural abscess . All medical records, laboratory data, hospital notes, radiological images, and operative reports were examined . Statistical analysis was performed by an independent statistician using the methods of least squares regression, analysis of variance, X2 , and rank correlation coefficient. Patients with less than 2 weeks of manifestations prior to admission were considered to have an acute presentation . Those who had longer symptomatology were classified as having a chronic course . Neurological deficit on admission was graded as severe if the patient was parapalegic, or moderate if the patient retained only antigravity strength . Ultimate outcome was considered good if recovery was complete or if the patient could ambulate without assistance ; fair if the neurological deficit improved and the patient was continent of bowel and blad0090-3079/92/35 .00
226
Surg Neurol 1992 ;38 :225-31
Nussbaum et al
Table 1 . Yearly Incidences
60 No . of
Year July July July July July July July July July July July July
Cases 1979-June 1980 1980-June 1981 1981-June 1982 1982-June 1983 1983-June 1984 1984-June 1985 1985-June 1986 1986-June 1987 1987-June 1988 1988-June 1989 1989-June 1990 1990-March 1991
Incidence per 10,000 Hospital Admissions
40
1
0 .53
SO
0 0
0 .0 0 .0
20
2 2 2
0 .95 0 .94 0 .93 0 .47 0 .48 0 .47
1
5 9 16
10
1079 1980 1981 1982 1983 1984 1986 1986 1987 1988 1989 logo
FISCAL YEAR
2 .26
4 .03 11 .31
der but enough weakness persisted to preclude independent ambulation ; and poor if deterioration occurred or if the patient was left incontinent or parapalegic . All magnetic resonance imaging (MRI) scans were obtained using a Siemens 1 .5-Tesla Magnetom MRI scanner (Iselin, NJ .) before and after intravenous injection of 0 .1 mmollkg gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) contrast material . All computed tomography (CT) images were obtained on a GE 9800 scanner (Milwaukee, Wisc .), with 3-mm contiguous sections, or on a Siemens DRH scanner (Iselin, NJ .) with 4-mm contiguous sections, immediately following the intrathecal instillation of contrast material (Omnipaque, Winthrop-Breon, New York) via lumbar or C 1-C2 interspace puncture . All erythrocyte sedimentation rates (ESR) were determined using the Westergren method .
Results Epidemiological Data Forty cases of spinal epidural abscess were reviewed (22 men and 18 women) . Ages ranged from 3 months to 72 years, with a mean age of 50 years . Yearly incidences, summarized in Table 1, reveal a statistically significant (least squares regression method, p = 0 .0195) increase in the incidence of epidural abscess after June 1988 (Figure 1) . Associated Clinical Findings We noted the following likely predisposing factors (Table 2) . Sixteen patients used illicit drugs intravenously, 13 had diabetes mellitus, and 12 had numerous medical disorders . Eleven patients reported having a history of recent (
225
Spinal Epidural Abscess : A Report of 40 Cases and Review Eric S . Nussbaum, M.D., Daniele Rigamonti, M .D., Harold Standiford, M .D., Yuji Numaguchi, M.D., Aizik L . Wolf, M.D., and Walker L . Robinson, M .D. Departments of Surgery, Division of Neurosurgery ; Diagnostic Radiology, Division of Neuroradiology ; and Internal Medicine, Division of Infectious Disease, University of Maryland Medical Systems, Baltimore Maryland
Nussbaum ES, Rigamonti D, Standiford H, Numaguchi Y, Wolf AL, Robinson WL. Spinal epidural abscess : a report of 40 cases and review . Surg Neurol 1992 ;38 :225-31 . Despite modern medical advances, the morbidity and mortality rates associated with spinal epidural abscess remain significant, and the diagnosis often is elusive . A retrospective study was undertaken to define better the incidence and clinical features of this infection, and to establish current diagnostic and therapeutic guidelines . Forty cases of spinal epidural abscess were encountered at our institution between July 1979 and March 1991 . All medical records and radiological images were reviewed . We report a significant increase in the incidence of epidural abscess after June 1988 (p = 0 .0195). Sixteen patients used drugs intravenously, and six had undergone spinal procedures. Twelve patients were misdiagnosed in various emergency rooms or clinics and discharged . Localized back pain, fever, and neurological deficit remained the typical clinical manifestations . Erythrocyte sedimentation rate was elevated uniformly when measured (21 cases) . Magnetic resonance imaging was diagnostic specifically in 23 of 24 instances . The majority of patients underwent surgical drainage, but five selected patients were managed nonoperatively . The highly variable presentation of spinal epidural abscess may confuse the diagnosis and delay indicated surgical intervention . Localized back pain in a febrile patient at significant risk for epidural abscess warrants erythrocyte sedimentation rate measurement . The presence of erythrocyte sedimentation rate elevation or evidence of spinal cord compression on physical examination are indications for immediate magnetic resonance imaging examination with contrast enhancement . Surgical drainage with sustained intravenous antibiotic treatment remains the cornerstone of therapy . Nonoperative management may be considered in selected cases . KEY WORDS : Spinal epidural abscess ; Spinal infection ; Spinal cord compression
Address reprint requests tar Daniele Rigamonti, M .D., University of Maryland Hospital, Division of Neurosurgery, Rm . NI2E27 22 South Greene Street, Baltimore, MD 21201 . Received November 18, 1991 ; accepted January 27, 1992 .
C 1992 by Elsevier Science Publishing Co ., Inc.
In 1948, Heusner studied 20 patients with spinal epidural abscess and emphasized the importance of early diagnosis [131 . Later, Baker and colleagues reviewed the clinical course and pathophysiology of epidural abscess [2] . Subsequent publications have stressed the potentially rapid deterioration and persistent high morbidity associated with this serious, localized infection [8,11,14,26,28,31,33-35] . Currently, the diagnosis of epidural abscess remains difficult, and the consequences of delayed intervention are still potentially devastating. However, since the time of the review by Baker et al, new imaging modalities have become available, more effective antimicrobial agents have been developed, and different risk factors have emerged as significant. Given these advances, our experience with spinal epidural abscess from 1979 to the present has been reviewed in an attempt to clarify issues such as incidence of epidural abscess and predisposing factors, diagnostic guidelines, treatment regimens, and outcome .
Methods and Materials From July 1, 1979, to April 1, 1991, the University of Maryland Medical Systems Hospitals admitted 245,643 patients . A careful review of patient records revealed 40 cases of documented spinal epidural abscess . All medical records, laboratory data, hospital notes, radiological images, and operative reports were examined . Statistical analysis was performed by an independent statistician using the methods of least squares regression, analysis of variance, X2 , and rank correlation coefficient. Patients with less than 2 weeks of manifestations prior to admission were considered to have an acute presentation . Those who had longer symptomatology were classified as having a chronic course . Neurological deficit on admission was graded as severe if the patient was parapalegic, or moderate if the patient retained only antigravity strength . Ultimate outcome was considered good if recovery was complete or if the patient could ambulate without assistance ; fair if the neurological deficit improved and the patient was continent of bowel and blad0090-3079/92/35 .00
226
Surg Neurol 1992 ;38 :225-31
Nussbaum et al
Table 1 . Yearly Incidences
60 No . of
Year July July July July July July July July July July July July
Cases 1979-June 1980 1980-June 1981 1981-June 1982 1982-June 1983 1983-June 1984 1984-June 1985 1985-June 1986 1986-June 1987 1987-June 1988 1988-June 1989 1989-June 1990 1990-March 1991
Incidence per 10,000 Hospital Admissions
40
1
0 .53
SO
0 0
0 .0 0 .0
20
2 2 2
0 .95 0 .94 0 .93 0 .47 0 .48 0 .47
1
5 9 16
10
1079 1980 1981 1982 1983 1984 1986 1986 1987 1988 1989 logo
FISCAL YEAR
2 .26
4 .03 11 .31
der but enough weakness persisted to preclude independent ambulation ; and poor if deterioration occurred or if the patient was left incontinent or parapalegic . All magnetic resonance imaging (MRI) scans were obtained using a Siemens 1 .5-Tesla Magnetom MRI scanner (Iselin, NJ .) before and after intravenous injection of 0 .1 mmollkg gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) contrast material . All computed tomography (CT) images were obtained on a GE 9800 scanner (Milwaukee, Wisc .), with 3-mm contiguous sections, or on a Siemens DRH scanner (Iselin, NJ .) with 4-mm contiguous sections, immediately following the intrathecal instillation of contrast material (Omnipaque, Winthrop-Breon, New York) via lumbar or C 1-C2 interspace puncture . All erythrocyte sedimentation rates (ESR) were determined using the Westergren method .
Results Epidemiological Data Forty cases of spinal epidural abscess were reviewed (22 men and 18 women) . Ages ranged from 3 months to 72 years, with a mean age of 50 years . Yearly incidences, summarized in Table 1, reveal a statistically significant (least squares regression method, p = 0 .0195) increase in the incidence of epidural abscess after June 1988 (Figure 1) . Associated Clinical Findings We noted the following likely predisposing factors (Table 2) . Sixteen patients used illicit drugs intravenously, 13 had diabetes mellitus, and 12 had numerous medical disorders . Eleven patients reported having a history of recent (
