Symptom-Based Diagnosis in Pediatrics (CHOP Morning Report) 1st Ed.

CASE 5-1

Two-Year-Old Boy



A 2-year-old boy presented to the emergency department for evaluation of back pain. Three days prior to admission he began complaining of abdominal pain, refused to eat lunch that day, and spent most of the afternoon watching television rather than playing outside with his siblings. At that time, he was taken to a nearby hospital for evaluation. On examination, he had mild diffuse abdominal tenderness but no rebound tenderness or involuntary guarding. Abdominal radiographs showed significant stool in the rectum and distal colon. He was diagnosed with constipation, given a glycerin suppository, and discharged after producing a moderate amount of stool.

On the day of admission, he returned to the hospital with persistent abdominal pain and a new complaint of low back pain. His oral intake had been poor over the past few days. There had been minimal response to a glycerin suppository earlier that day. He also seemed particularly uncomfortable while his diaper was being changed. There was no fever, cough, hematemesis, hematochezia, dysuria, or urinary frequency. There were no ill contacts and no known trauma. The only pet was an elderly dog that had been euthanized earlier in the week.


Tympanostomy tubes had been placed at 15 months of age for recurrent otitis media. He had only one episode of otitis media after the tubes were placed. He did not have a prior history of constipation. He was not taking any medications on a regular basis. His family history was remarkable for a paternal uncle who had a myocardial infarction at 55 years of age.


T 38.9°C; HR 130 bpm; RR 36/min; BP 115/55 mmHg; SpO2 99% in room air

Weight 18.0 kg (>95th percentile)

The child appeared uncomfortable and refused to stand. The eyes, nose, and oropharynx were clear. The neck was supple. The abdomen was mildly distended and diffusely tender, particularly in the right lower quadrant. However, there was no rebound tenderness or involuntary guarding. There was no costovertebral angle tenderness. There was discomfort with passive flexion of the right hip. There was mild edema and tenderness to percussion along the right paraspinus muscle at the level of the L1 vertebrae. There was no kyphosis, scoliosis, or abnormal lordosis. There were no apparent sensory or motor neurologic deficits, though the degree of back and abdominal pain made assessment of muscle strength in the lower extremities difficult. There was no muscle atrophy. Rectal tone was normal. The deep tendon reflexes were symmetric and appropriately brisk. The remainder of the examination was normal.


Complete blood count revealed the following: 19 700 white blood cells/mm3 (67% segmented neutrophils, 29% lymphocytes, and 3% monocytes); hemoglobin, 11.4 g/dL; and platelets, 390 000/mm3. Serum electrolytes were remarkable for a bicarbonate level of 19 mEq/L and for blood urea nitrogen and creatinine levels of 7 mg/dL and 0.3 mg/dL, respectively. Urinalysis revealed a specific gravity of 1.020 and 3+ ketones but normal microscopy. Serum albumin and transaminases were normal. C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) were elevated at 7.9 mg/dL and 65 mm/h, respectively. Abdominal obstruction series revealed scattered air fluid levels and a small amount of stool in the rectum.


Magnetic resonance imaging (MRI) of the spine localized the abnormality (Figure 5-1) and the definitive diagnosis was made by an interventional study.


FIGURE 5-1. Magnetic resonance of the spine.



Back pain is a relatively common complaint among children though fewer than 2% of children with back pain require specific medical evaluation. In a young child, neoplastic, infectious, and inflammatory disorders should be considered. Traumatic injury is usually clear based on the history prior to presentation. Malignant causes may be primary or metastatic and include osteoid osteoma, neuroblastoma, Wilms tumor, and leukemia. Infectious or inflammatory causes include pyelonephritis, vertebral osteomyelitis, spinal epidural abscess, and pyomyositis. Diskitis in children usually involves the lower thoracic or lumbar spine. The presence of fever, if related to the back pain, makes diskitis less likely. Local tenderness and elevated CRP and ESR can be seen with many infectious and neoplastic causes. MRI of the spine readily differentiates diskitis, vertebral osteomyelitis, and spinal epidural abscess. Rheumatologic conditions include systemic juvenile rheumatoid arthritis and ankylosing spondylitis. Mechanical disorders, such as muscle strains and intervertebral disk herniation, are less likely in children, although spondylolysis and spondylolisthesis do present in this age group. The absence of neurologic findings, while reassuring, does not exclude any of the above entities.


MRI of the spine revealed an abnormal heterogeneous enhancing mass (Figure 5-1, arrows) in the epidural space at the L1-L3 vertebral level. There was associated compression of the thecal sac. Urine homovanillic acid and vanillylmandelic acid levels were normal, making neuroblastoma less likely. Gram-stain of purulent material drained during biopsy of the mass revealed many white blood cells and Gram-positive cocci. Group A Streptococcussubsequently grew from culture. The diagnosis was spinal epidural abscess due to group A Streptococcus. The demise of the pet dog did not appear to be related to this patient’s diagnosis.


Spinal epidural abscesses occur rarely in children with one series reporting an incidence of 0.6 per 10 000 hospital admissions over a 15-year period at a free standing children’s hospital. Most patients are previously healthy but predisposing risk factors include sickle cell disease (SCD), hematologic malignancy, and spinal surgery. Spinal epidural abscess occasionally complicates serial lumbar punctures and varicella infection. In adult patients, additional risk factors include trauma and invasive procedures (i.e., spinal aneasthesia). Staphylococcus aureus causes more than two-thirds of cases; additional pathogens have been reported (Table 5-3). With the emergence of community acquired methicillin-resistant Staphylococcus aureus (MRSA), the incidence of MRSA epidural infections has increased over recent years.

TABLE 5-3. Etiologies of epidural abscesses.


The infection is usually acquired by hematogenous spread and occasionally by direct extension from an adjacent site of infection. Associated osteomyelitis is present in approximately 50% of cases. In one case series, 7 of 8 children with spinal epidural abscess had an associated psoas or para-spinal abscess.


Most children develop fever early during the course of infection. Common presenting complaints include back pain, limp, and refusal to walk. Hip pain is an unusual presenting complaint though it may be difficult to differentiate back from hip pain in an ill and irritable child. Depending on the level of involvement, progression of infection can cause spinal cord compression, muscle weakness, and bowel and bladder incontinence followed by paralysis. On examination, there may be tenderness over the vertebra or paraspinal tissues. Some children develop protective paraspinal muscle spasm. There may also be loss of normal curvature of the spine (usually decreased lumbar lordosis) and limited lumbosacral mobility. Abdominal pain is relatively common and can indicate radicular pain or associated psoas abscess.


Spinal epidural abscesses can present with a range of clinical and laboratory findings and a high level of suspicion is required to make the diagnosis early in the course of infection. Surgical aspiration should always be performed since identification of a specific pathogen permits optimal antibiotic selection. Other studies may increase the level of suspicion for spinal epidural abscess.

Complete blood count. The complete blood count typically reveals the nonspecific findings of acute infection. The peripheral white blood cell count is elevated in approximately 50% of cases. There may be a predominance of neutrophils or an increased percentage of immature polymorphonuclear cells. Thrombocytosis may be present.

CRP and ESR. These markers of inflammation are usually elevated, especially when there is an associated vertebral osteomyelitis. These markers of inflammation may also be elevated in noninfectious conditions such as malignancy. CRP and ESR have also been used to assess response to antibiotic therapy. The timing of their normalization has not been studied with epidural abscesses but likely parallels their trends in osteomyelitis where the CRP peaks within the first 2 days of treatment and returns to normal within 7-10 days and the ESR peaks within the first 5-7 days of treatment and returns to normal within 4 weeks.

Blood culture. Organisms are isolated from blood culture in approximately 10% of cases. When positive, the blood culture is invaluable in guiding specific antibiotic therapy.

Spine radiographs. Radiographs of the spine exclude other causes of back pain. Associated vertebral osteomyelitis may be evident in children with a prolonged duration of symptoms.

Spine MRI. MRI of the spine demonstrates the abscess, though definitive diagnosis requires biopsy. MRI reveals concomitant vertebral osteomyelitis in 20% to 50% of cases.

Tuberculin skin testing. Tuberculin skin testing should be performed if a bacterial organism has not been isolated from blood or abscess culture since Mycobacterium tuberculosis can cause spinal epidural abscesses.

Other studies. At the time of diagnostic biopsy, specimens should be sent for stains and cultures of aerobic and anaerobic bacteria, fungi, and myco-bacteria. Radionuclide bone scans to detect osteomyelitis at sites distant from the abscess should be considered in cases where the abscess occurred as a consequence of hematogenous seeding. Cerebrospinal fluid (CSF) abnormalities are common with spinal epidural abscesses. In one case series, 33 (78%) of 42 children with spinal epidural abscess had findings consistent with meningeal infection (mild to moderate pleocytosis or hypoglycoracchia). In 12%, elevated CSF protein was the only CSF abnormality. Examination of the CSF was completely normal in 10% of children with spinal epidural abscess. Lumbar puncture should not be performed when the abscess is located in the lumbar region.


Standard management of epidural abscesses includes antibiotic therapy and surgical drainage. Sporadic cases reported in the literature have been treated with antibiotics alone. Candidates for antibiotic therapy without surgical drainage may include patients without neurologic deficits and those with numerous abscesses that would be technically difficult to drain. In those children treated with antibiotics without surgical drainage, diagnostic surgical aspiration to identify the infecting organism should be strongly considered. This decision is usually made in consultation with infectious diseases and neurosurgical colleagues. The empiric antibiotic regimen should include agents with activity against Staphylococcus aureus such as oxacillin or vancomycin. Vancomycin should be the initial antibiotic when (1) MRSA accounts for more than 10% to 15% of local S. aureus isolates, (2) a household member works in a nursing home or other facility with high rates of MRSA colonization, and (3) the patient lives with someone known to be colonized with MRSA. Cefotaxime and metronidazole should be added if Gram-negative or anaerobic organisms are suspected. With increasing rates of Gram-negative resistance to third generation cephalosporins, many practitioners now recommend cefepime rather than cefotaxime. Ultimate antibiotic selection depends on results of blood and abscess culture. The duration of antibiotic treatment is usually determined by a combination of clinical (e.g., improved pain and function), laboratory (e.g., normalization of ESR and C-reactive protein levels), and radiologic imaging (e.g., resolved epidural fluid collection on MRI) improvement, but 6 weeks is usually the minimal duration of therapy.

Mortality rates for adults with spinal epidural abscesses range from 5% to 25%. Mortality rates are substantially lower in children. There were no deaths among the 34 children reviewed in one series. Approximately 75% to 85% of children treated for spinal epidural abscess will have normal neurologic function at the completion of therapy. Risk factors for persistent deficits include patients with multiple medical problems, previous spinal surgery, and severe neurologic deficit at presentation.


1. Auletta JJ, John CC. Spinal epidural abscesses in children: a 15-year experience and review of the literature. Clin Infect Dis. 2001;32:9-16.

2. Grewal S, Hocking G, Wildsmith JA. Epidural abscesses. Bri J. Anaesth. 2006;96:292-302.

3. Darouiche RO. Spinal epidural abscess. N Engl J Med. 2006;355:2012-2020.

4. Bair-Merritt MH, Chung C, Collier A. Spinal epidural abscess in a young child. Pediatrics. 2000;106:e39. (

5. Yogev R. Focal suppurative infections of the central nervous system. In: Long SS, Pickering LK, Prober CG, eds. Principles and Practice of Pediatric Infectious Diseases. 3rd ed. Philadelphia: Churchill Livingstone; 2008:324-335.

6. Rubin G, Michowiz DS, Ashkenasi A, Tadmor R, Rappaport H. Spinal epidural abscess in the pediatric age group: case report and review of the literature. Pediatr Infect Dis J. 1993;12:1007-1011.

7. Tunkel AR. Subdural empyema, epidural abscess, and suppurative intracranial thrombophlebitis. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases. 7th ed. Philadelphia: Churchill Livingstone; 2010:1279-1287.