The Back Pain Differential

1 Comment
You’re working at a teaching hospital ED in Houston one evening when one of your residents presents a case of back pain. The patient is a 57-year-old African-American female with a history of Hepatitis C diagnosed in 2000, questionable cirrhosis, macrocytic anemia, GERD, and chronic back pain for five years. Apparently the patient has close follow-up with her PCP and has suffered back problems for some time, and treatment with Flexeril and Darvocet was no longer relieving her pain. She started seeing a pain specialist one month ago and had received epidural nerve block injections to the lumbar spine every two weeks. The patient had her third and last epidural injection eight days prior to her presentation to the ED. Today she comes in complaining of worsening back pain that is sharp in nature, throbbing, 10/10 in severity with radiation down her legs intermittently. The patient also complained of progressively worsening leg weakness to the point that she could no longer ambulate, and numbness and tingling of her left lower extremity. Her right leg also felt weaker than usual and she had shooting pains going down her leg occasionally. Associated symptoms included a temperature of 100.6°F the night before, night sweats for one week duration, difficulty walking, constipation, and decreased appetite. She also reported difficulty lying on her back due to pain and tenderness over her low back when lying supine.

Her medications include only Nexium for GERD, Lasix 20mg po daily as needed for leg edema, Flexeril and Darvocet prn for back pain. Past surgical history was significant for hysterectomy in 1973 for uterine fibroids and a right knee arthroscopy in 1989. The patient lives in Houston with her husband, was retired, and denied any tobacco, alcohol, or illicit drug use. She had no travel history, no pets, nor any sick contacts.

When you see the patient, she is in moderate distress and appears unable to find a comfortable position to lie on the stretcher even after receiving a dose of IV Morphine. Her pulse is 90, she is afebrile (T=98.7° F), blood pressure 144/72, breathing comfortable at 14 breaths per minute and O2 sats of 99% on room air. Cardiac exam is normal with no murmurs noted. Abdominal exam reveals moderate ascites with shifting dullness, no tenderness on palpation, bowel sounds present, no rebound, no guarding, no hepatosplenomegaly. Her extremities have 1+ pitting edema to her knees bilaterally, no clubbing or cyanosis is noted, and pedal pulses are 2+ equal and symmetric bilaterally. Her neurologic exam reveals patellar tendon and ankle reflex of 3+ in the right lower extremity 2+ in the left lower extremity. Strength is 3/5 in the right lower extremity and 4/5 in the left lower extremity. Straight leg lift of left leg causes shooting back pain down to great toe. Sensation including sharp vs. dull, proprioception, and hot vs. cold are intact in upper and lower extremities bilaterally. Examination of her back shows no erythema, no lesions, nor fluctuance, but her lower lumbar spine (L4-L5) and upper sacral region is tender to light palpation. Rectal exam shows no masses or lesions and normal sphincter tone.

The patient’s labs are unremarkable except for WBC=11.8 K/uL with 72%neutrophils and 14% monocytes, Platelet count=87K/uL, with hemoglobin=11.7 g/dL, hematocrit=34.5%, and MCV=123fL. Pt’s HIV is negative. Sedimentation rate=6 mm/Hr.


In the setting of fevers, back pain, and progressive leg weakness, the patient underwent an MRI for further evaluation for an epidural abscess. Neurosurgery was consulted while awaiting the MRI. Two sets of blood cultures were drawn, and the patient was given one dose of Vancomycin 1 gram IV, Metronidazole 500mg IV and Ceftazidine 2 grams IV empirically.




The patient’s MRI images appear. What do you see? What do you do next?

Dx: Vertebral Osteomyelitis

MRI results: “Abnormal enhancement in the L5 and S1 vertebral bodies, abnormal enhancement and degenerative changes in the facets, abnormal enhancement prevertebral, paraspinal and epidural soft tissues without abscess formation. All of these findings are suspicious and concerning for discitis/osteomyelitis. Mild bilateral foraminal stenosis due to mild degenerative disc disease and facet arthrosis. No spinal canal stenosis. Otherwise, no lumbar abnormalities. “


With fever, back pain, and neurologic deficits, the main diagnosis to consider is spinal epidural abscess. However, that triad is only 7.9% sensitive for epidural abscess. Thus, if you wait for it, you’re going to miss an epidural abscess or two. While considering epidural abscess, you should also be thinking about another zebra, vertebral osteomyelitis.


Spinal epidural abscess (SEA) has an estimated incidence of 1-2 per 10,000. Most patients with SEA have predisposing factors including immunocompromised states (including diabetes), underlying spinal abnormality, or a potential source of infection. Hematogenous spread occurs in about half of patients with direct inoculation occurring in 1/3. The classic triad is back pain (3/4 of patients), fever (approx ½ of patients), and neurologic symptoms (approx 1/3 of patients), although very few patients present with all three. The progression of SEA has 4 distinct stages that include (1) back pain, (2) nerve root pain, (3) motor weakness, sensory deficit, bowel/bladder dysfunction, (4) paralysis. Unfortunately, the progression of symptoms is highly variable and worsening neurologic deficits or paralysis can progress within hours.

In contrast, vertebral osteomyelitis, a term often used interchangeably with discitis, refers to infection or inflammation of the vertebral bones, which usually spreads to the adjacent disc space. Occasionally, the process can start in the disc space and move to bone, especially after surgery or pain injections. Most patients are over 50 years old and the incidence is estimated to be approximately 1:450,000 people. Spread can occur by either contiguous spread from adjacent tissue or by direct inoculation, but may also occur via hematogenous routes. In fact, hematogenous spread is, by far, the most common route of infection. Furthermore, facet joint or epidural injections are rare causes of discitis. The most prominent symptom of discitis is back pain that has usually persisted for greater than one month. The pain is often worse at night and may be relieved partially by bed rest. Fever may accompany the symptoms, but this is not the rule. Other common symptoms include an exaggerated lumbar lordosis and difficulty/painful walking with an otherwise normal neurologic exam. Symptoms mimicking epidural abscess may be present if the infection spreads posteriorly.

Laboratory studies include a CBC for leukocytosis (only elevated in 2/3 of SEA patients), ESR (elevated in approx 80% of discitis patients but almost uniformly in SEA), and CRP.7 Blood cultures are also obtained and may be positive in up to 70% of patients with discitis. In SEA, blood cultures are almost always positive in patients with positive CSF cultures, but the exact sensitivity is unknown. The patient with either diagnosis may also need an evaluation for endocarditis if there are positive blood cultures, infection from gram-positive organisms, or predisposing structural heart disease.

Magnetic resonance imaging (MRI) is the most sensitive imaging study for detecting spinal epidural abscess and vertebral osteomyelitis.Abnormalities on MRI will likely precede changes on plain films and can pick up the early erosive changes that may be missed by CT scan. Needle biopsy is usually necessary to confirm the diagnosis if discitis.

Empiric antibiotics should be started before diagnosis is confirmed when there is concern for epidural abscess. Pathogens include S. aureus (including MRSA), enteric gram-negative bacilli, streptococci species, pseudomonas and Candida species (the latter two are most associated with intravascular drug abuse). Broad-spectrum coverage is recommended. One such regimen is vancomycin and ceftazidime as was used in this patient.

For discitis, most patients will respond to antibiotics, and surgery is rarely necessary. The prognosis of discitis seems to be determined by the time of delay to diagnosis and presence of neurologic symptoms at time of diagnosis. A recent study by Dimar and colleagues estimates the mortality at approx 5%. Morbidity (recurrence and relapse rates) are difficult to determine, but are in the range of 0-30%. This is in stark contrast to SEA, in which surgical decompression is the standard of care, and neurologic function at discharge most often reflects the patient’s deficits at the time of diagnosis. Irreversible paralysis occurs in 4-22% with delay of diagnosis being the biggest cause. Mortality for SEA is only estimated at 5%, usually from sepsis or meningitis.

This patient presented with several signs and symptoms concerning for both spinal epidural abscess and vertebral osteomyelitis. As mentioned above, discitis can present similarly to SEA when the posterior vertebral bodies are affected, making them virtually indistinguishable. Fortunately, the diagnostic work-up for both are the same. In our cost-conscious, yet litigation heavy society, this patient may present an additional diagnostic problem to those without ready access to MRI or neurosurgical consultants. Although this patient’s admitting diagnosis was discitis, the primary diagnosis to rule out at the time of presentation was epidural abscess. SEA cannot be ruled out without an MRI. Plain CT has poor sensitivity for epidural abscess. CT with intrathecal gadolinium can also be useful, but is not as sensitive as MRI and is technically difficult to perform. Myelography has become obsolete. To avoid permanent neurologic deficits, there should be as few delays as possible for prompt diagnosis and treatment. Delays in care to confirm the diagnosis could cost the patient their ability to walk or control their bowel/bladder! In the community, if an MRI is available, a call should be placed to the on-call neurosurgeon while the patient receives antibiotics to discuss the best treatment plan and make the decision to transfer the patient earlier rather than later. If an MRI is unavailable, then the patient should be transferred to an appropriate facility as soon as possible.

Conclusion: Neurosurgery was consulted in ED for further management. After a bone biopsy was obtained via interventional radiology, the patient was empirically started on Bactrim, Ciprofloxacin and Rifampin. Tissue biopsy gram stain, AFB, fungal, and aerobic/anaerobic cultures did not reveal any organisms. Infectious diseases was consulted for further management, and after reviewing her cultures and pathology, antibiotics were discontinued, She received physical therapy to assist with ambulation and was discharged to a rehabilitation facility with close follow-up. Although the spectrum of disease for discitis ranges from a non-infectious inflammatory process to the most severe, osteomyelitis, the initial presentation for patients along that spectrum provides little diagnostic guidance as to how severe the cause may be and as to whether or not it will ultimately be caused by an infectious etiology. Despite the fact that our patient was one of the lucky ones, experiencing a self-limited inflammatory process, her presentation was indistinguishable from cases with more severe pathology such as vertebral osteomyelitis or even spinal epidural abscess. For the emergency physician, the choices are clear. When we’re considering the diagnoses of discitis or spinal epidural abscess, have a low threshold for pulling the trigger on MRI, antibiotics and early neurosurgical consultation.

1 Comment

Leave A Reply