Spinal epidural abscess

[Angry Birds]

how do I make sure not to miss this given I see so many back pain patients? I feel like this diagnosis was not emphasized at all in my residency and I only really even considered it if 1) IVDA or Diabetes, 2) fever, 3) point tenderness, or 4) neuro deficits. But none of these need to be present, so any advice?

---

Members don't see this ad.

 

[Doctor J]

considered it if 1) IVDA or Diabetes, 2) fever, 3) point tenderness, or 4) neuro deficits.

That's a pretty good list. In general if you ask about and document absence of any neuro deficits or fever and your return instructions warn against worsening symptoms such as neuro deficits and fever, even if you miss this diagnosis, you're practicing within the standard of care at most places I can think of.

Back pain is really common. Emergent MRIs for back pain are not.

I also make sure to walk my back pains. If they can't walk they can't go home. You'll catch some this way.

 

[Zebra Hunter]

so your question is 'how do I make the diagnosis of spinal epidural abscess if there are no red flags?' ...? The answer is you don't. That is unless you want to inappropriately scan hundreds to thousands of patients over your career.

 

[WilcoWorld]

how do I make sure not to miss this given I see so many back pain patients? I feel like this diagnosis was not emphasized at all in my residency and I only really even considered it if 1) IVDA or Diabetes, 2) fever, 3) point tenderness, or 4) neuro deficits. But none of these need to be present, so any advice?

You're doing it right. To add a safety net, when I'm dc'ing back pain patients I tell them to come back if they develop a fever or a deficit, or if the pain gets a lot worse.

 

[MSmentor018]

5. spinal injections
6. any implanted devices
7. long term steroid use
8. immunosuppressed
9. coagulopathic
10. cancer

I have a standard approach to all these patients. detailed physical exam including rectal, post void residual. patellar reflexes, saddle anesthesia. I do the post void w bedside u/s.
if anything on the H+P is positive, they get a work up and crp/esr to labs . it's not an exact science but if those are neg, chance of abscess are pretty low. then it's +/- ct/mri

my population are mostly old with money. they all have pain mgt, acupuncture, naturalpaths, chiropractics, vetebralplasty 3-4 times, uses specialist for their meds ("because they've had more training") and webmd is their primary . hidden in there are a few heroin using grandmas. compression fx are a dime a dozen, couple of AAA >5cm. about 7 cord/abscess/hematoma cases per yr. and that's what I see.

 

[Angry Birds]

Thank you very much, guys! In addition to the list 5-10, one thing I'll change to my practice is discharge instructions. In any case on all back pain patients I cover myself by urging an urgent outpatient MRI by your PCP if symptoms persist or worsen. Probably an expensive thing for society but does cover me.

 

[Arcan57]

Thank you very much, guys! In addition to the list 5-10, one thing I'll change to my practice is discharge instructions. In any case on all back pain patients I cover myself by urging an urgent outpatient MRI by your PCP if symptoms persist or worsen. Probably an expensive thing for society but does cover me.

That is a jerk move. Back pain is going to persist and in many cases it will worsen from initial visit. The natural course of spinal epidural abscesses is multiple healthcare contacts over the course of weeks prior to diagnosis. If you're telling every acute lumbago with 1st presentation they need an MRI, you're just reinforcing the message that low back pain is a surgical disease for which there is fast/effective treatment. It may initially cover you for the particular encounter. However, it's going to create significantly more ED visits for back pain as they either 1) don't have PCPs, 2) the next appointment for their PCP is 4-6 weeks away, or 3) they can't afford the copay/insurance won't cover outpt MRI. Since you've now created the expectation that an MRI is needed, the next doc is going to have to deal with a nightmare as they have a patient that's still having pain 2 weeks later, has no neuro deficits, and was told by one of the EP's own colleagues that they would need an MRI to prevent becoming paralyzed or impotent.

Having just run this work-up yesterday, even with an in-house MRI tech and the table open it still took about 4 hrs to get an MRI w/ & w/o T&L spine with a nurse having to go down to MRI 3 different times to administer pain meds due to pt motion.

Think of epidural abscess like aortic dissection or other "low incidence of disease/high incidence of symptomatology - high morbidity conditions". Think about it in every patient, document an exam that's demonstrates that you thought about it, get definitive imaging on patients that you are concerned about, don't image people that you feel are low risk, and give appropriate discharge instructions such that if you're wrong there's a chance the patient will have another healthcare contact prior to decompensation.

 

[deleted206819]

This is the Epic smartphrase I use:

Patient currently denies any recent history of fever, neurologic symptoms, urinary incontinence or retention. Patient denies any history of cancer, IVDU, AAA, AVM, spinal surgery or recent trauma, steroid use.

I have considered the possibility of SEA (spinal epidural abscess), the patient has no features that places them at increased risk of SEA and they have a normal neurological examination in the emergency department. The pre-test probability of SEA is sufficiently low that they do no need any further workup emergently.

Physical examination shows present and equal dorsalis pedis pulses, maintains dorsiflexion of great toes, no saddle paresthesia, good rectal tone, and negative straight leg raise.

I have low suspicion for cauda equina, renal etiology, AVM or aortic aneurysm/dissection, fracture, tumor, or infection.

 

[neusu]

Even if it is there, it is not operative until there are deficits.

 

[RustedFox]

This is the Epic smartphrase I use:

Patient currently denies any recent history of fever, neurologic symptoms, urinary incontinence or retention. Patient denies any history of cancer, IVDU, AAA, AVM, spinal surgery or recent trauma, steroid use.

I have considered the possibility of SEA (spinal epidural abscess), the patient has no features that places them at increased risk of SEA and they have a normal neurological examination in the emergency department. The pre-test probability of SEA is sufficiently low that they do no need any further workup emergently.

Physical examination shows present and equal dorsalis pedis pulses, maintains dorsiflexion of great toes, no saddle paresthesia, good rectal tone, and negative straight leg raise.

I have low suspicion for cauda equina, renal etiology, AVM or aortic aneurysm/dissection, fracture, tumor, or infection.

I like this, a lot.

I'm a-gonna post my smartphrase when I get ahold of it next (I'm off tonight and tomorrow, so fuh-get it.) Together, we can build the best smartphrase, ever.

 

[WilcoWorld]

This is the Epic smartphrase I use:

Patient currently denies any recent history of fever, neurologic symptoms, urinary incontinence or retention. Patient denies any history of cancer, IVDU, AAA, AVM, spinal surgery or recent trauma, steroid use.

I have considered the possibility of SEA (spinal epidural abscess), the patient has no features that places them at increased risk of SEA and they have a normal neurological examination in the emergency department. The pre-test probability of SEA is sufficiently low that they do no need any further workup emergently.

Physical examination shows present and equal dorsalis pedis pulses, maintains dorsiflexion of great toes, no saddle paresthesia, good rectal tone, and negative straight leg raise.

I have low suspicion for cauda equina, renal etiology, AVM or aortic aneurysm/dissection, fracture, tumor, or infection.

This is great. I would also add "no tag-backs."

 

[GatorCHOMPions]

I think this is a fairly useful algorithm. It has a small N, but they use most of the red flags mentioned with the addition of ESR/CRP.

https://umem.org/files/uploads/1108222010_JC972011B.pdf

 

[TooMuchResearch]

I think this is a fairly useful algorithm. It has a small N, but they use most of the red flags mentioned with the addition of ESR/CRP.

https://umem.org/files/uploads/1108222010_JC972011B.pdf

Link didn't work for me for some reason.

Are they labbing up all their back pain patients?

Ain't nobody got time for that.


Also, no one I know is routinely doing a rectal exam for back pain.

 

[GatorCHOMPions]

Link didn't work for me for some reason.

Are they labbing up all their back pain patients?

Ain't nobody got time for that.


Also, no one I know is routinely doing a rectal exam for back pain.

The link above was to the full paper and still works for me, but I have also attached it as a pdf. The pdf contains a nicely laid out algorithm. If you're lazy and just want the abstract, here is the link: https://www.ncbi.nlm.nih.gov/pubmed/21417700

Basically if you lack a progressive neuro deficit, but still have some risk factors or fever and esr/crp are low, you can discharge with follow up. If you lack fever, radicular pain, or risk factors, you simply discharge (this is probably what most of us do currently, anyway). It is labeled as a "clinical decision guideline" by the Journal of Neurosurgery which is the official journal of the Academic Association of Neurological Surgeons, so it is pretty legit.

 

[turkeyjerky]

Link didn't work for me for some reason.

Are they labbing up all their back pain patients?

Ain't nobody got time for that.


Also, no one I know is routinely doing a rectal exam for back pain.

Basically they're labbing anyone with "spine" pain, radicular pain, neuro deficits, or risk factors. Which, in my patient population, would be pretty much everyone. ("Yah doc I got tha sugar")

 

[TooMuchResearch]

Basically they're labbing anyone with "spine" pain, radicular pain, neuro deficits, or risk factors. Which, in my patient population, would be pretty much everyone. ("Yah doc I got tha sugar")

Oh, we don't do that.

 

[gman33]

This is the Epic smartphrase I use:

Patient currently denies any recent history of fever, neurologic symptoms, urinary incontinence or retention. Patient denies any history of cancer, IVDU, AAA, AVM, spinal surgery or recent trauma, steroid use.

I have considered the possibility of SEA (spinal epidural abscess), the patient has no features that places them at increased risk of SEA and they have a normal neurological examination in the emergency department. The pre-test probability of SEA is sufficiently low that they do no need any further workup emergently.

Physical examination shows present and equal dorsalis pedis pulses, maintains dorsiflexion of great toes, no saddle paresthesia, good rectal tone, and negative straight leg raise.

I have low suspicion for cauda equina, renal etiology, AVM or aortic aneurysm/dissection, fracture, tumor, or infection.

This is great. Hopefully you document all the exam findings in the chart as well.

Problem is that if there is a bad outcome, we are still getting sued.

 

[goodoldalky]

The link above was to the full paper and still works for me, but I have also attached it as a pdf. The pdf contains a nicely laid out algorithm. If you're lazy and just want the abstract, here is the link: https://www.ncbi.nlm.nih.gov/pubmed/21417700

Basically if you lack a progressive neuro deficit, but still have some risk factors or fever and esr/crp are low, you can discharge with follow up. If you lack fever, radicular pain, or risk factors, you simply discharge (this is probably what most of us do currently, anyway). It is labeled as a "clinical decision guideline" by the Journal of Neurosurgery which is the official journal of the Academic Association of Neurological Surgeons, so it is pretty legit.

Yeah and the post-intervention delayed diagnosis rate was still 10% based on their algorithm. This diagnosis is a b^&ch and you'll for sure see it a couple times. Out of like 48 million back pain patients. You have to get a little bit lucky and make a habit out of specifically asking people about IV drug abuse, and then maybe still miss it.

 

[BAM!]

... but there are sooooo many diabetics with back pain.
... and soooo many back pain with lumbar radiculopathy.
Surely we're not expected to MR all these, right?
In my book, you need a positive neurologic physical exam finding OR a higher risk factor, like diabetes with fever, or HIV, recent surgery, etc.

I can't imagine doing an MR on the rhematoid arthritis back pain because they're on methotrexate.

 

[BAM!]

I like this, a lot.

I'm a-gonna post my smartphrase when I get ahold of it next (I'm off tonight and tomorrow, so fuh-get it.) Together, we can build the best smartphrase, ever.

I like this idea. Together we should build a smart-phrase library for all sorts of complaints and sticky thread it. Though if it got too popular, I'm sure the lawyers would find it and somehow delegitimize your charting during a trial.