Thank you.

Thank you all for hanging in there during these stressful times. And thanks to your families, for understanding the difficult work you all do. And finally, thank you for helping grow this project to make us all better and safer clinicians. If you find MMI's stories useful, please continue to share it with your colleagues.

Great catch of a spinal epidural abscess (SEA)

Atypical history, no risk factors, first ED visit

Facts: An adult female is unable to sleep due to severe back px, spasms and swelling for 3 days. She drives herself to the ED in the pre-dawn hours. There she requires a WC to triage. Px is worse in the L lumbar region with px, numbness and cold sensation in her L leg. She also describes intermittent inability to move her L leg. ROS is negative for fever, N/V, GU sx, or recent infection. Pertinent PH is positive for agoraphobia, bipolar disorder and PTSD post combat. PH is negative for diabetes, alcoholism, IVDU, immune compromise or surgical hardware; she later acknowledges a dental problem of uncertain importance. OTC px meds have not helped. SH/FH is positive for mother's severe back pain and death from ALS. She perseverates, answers some questions inappropriately and is a generally poor historian. VS are normal. She has tenderness only in the L lumbar area. Her neuro exam shows no leg weakness but absent L patellar and ankle reflexes. She is given Toradol IM and her care is appropriately transferred to the day doc when the overnight doc cannot get her to consent to an MRI. Labs are drawn and are all normal except for a WBC of 16,200 w/ 83% polys. She is given lorazepam 4 hours post arrival and noted to have a temp of 38.1 7 hours post arrival. That rises to 39.1 only 30 minutes later. An SEA is considered, she is started on vancomycin and given MS for pain. She also finally agrees to a non-contrast thoracic/lumbar MRI which shows a presumptive SEA from T-2-T11, confirmed by contrast MRI. She is taken emergently to the OR for decompression 12 hours after arrival.
Plaintiff: None. SEA was considered, found and appropriately treated.
Defense: Excellent care despite patient being atypical, difficult and uncooperative.
Result: Pt. is moving both legs the following day and is walking without assistance 4 weeks later.
Takeaways:
* SEA was considered by both docs. This is key to making the diagnosis and should be the first thing to consider when seeing one of the many ED patients with back pain. A minimum level of suspicion is all that is needed to avoid missing a life-changing diagnosis. H & P can usually eliminate if from the differential.
* When the going gets tough, the tough keep going. Persistence matters. Don't give up on difficult patients.
* Consider transferring care to or consulting with a colleague when the picture is unclear and you are stymied.
* The "classic triad" was not present on arrival; there was no fever, and neuro sx were inconsistent.
* The elevated WBC was a good clue.
* A CRP/ESR could have been done, but the decision was made that she would need the MRI regardless. Either a CRP or ESR would likely have been markedly elevated.
Reference: Errors in Diagnosis of Spinal Epidural Abscesses in the Era of Electronic Health Records. Bhise V, Meyer A, Singh H, Wei L. Amer J Med 130(8), March 2017. Abstract and free full-text download available here.
[Editor's Note: The above reference is excellent but says little about a common “red flag”: escalating ED or clinic visits for back pain over days to weeks. The dx is easy to make if one thinks about it. It’s ruled out with 99% sensitivity if a CRP or ESR is normal - and you'll have proof that you included SEA in your differential. /cp]

One more for the road. Don't be next.

Too many SEA's still being missed.

Facts: Three ED visits for back pain in 12 hours. Note the biases and delays.
ED Visit 1: A female in her early 30's presents to the ED with 6/10 mid-back px radiating to her chest. She relates the onset to lifting her daughter’s carseat out of the car 2 days prior. Her ROS includes slight SOB, no cough, slight nausea and no fever. A CXR is ordered and suggests a subtle pneumonia or infarct. A chest CT is then done and finds only “airspace disease” and subtle atelectasis. She is given Dilaudid 1 mg, Zofran 4 mg, Toradol 30 mg and Ativan 0.5mg IV while in the ED, and discharged with a diagnosis of “pneumonia” with prescriptions for azithromycin and Vicodin.
ED Visit 2: She returns to the ED 11 hours later with 6/10 back px and bilateral leg weakness, numbness, tingling and pain radiating to her chest. She says that the new symptoms started shortly after taking her first dose of azithromycin. On this visit absence of incontinence and fever is documented but so is a history of bipolar disorder. On exam she is noted to be "anxious" and able to raise both legs against gravity. Sensation and reflexes are not documented. Physician notes “patient asking to go home.” She is switched from azithromycin to doxycycline and discharged with a dx of “paresthesias.” The discharge nurse records “Normal sensation from head to lower rib margin. Px 5/10. Pt reports MD will be discharging her home.”
ED Visit 3: She returns 45 minutes later by EMS after falling at home and is now unable to move her legs except to withdraw to pain. Babinskis are positive and a bladder scan shows 500 cc of urine. An LP is done; CSF protein is 335. An MRI is ordered but not done for 3 hours, eventually revealing a spinal epidural abscess extending caudad from T-6. A neurosurgeon at the system's "mother ship" hospital accepts patient in transfer, but since no bed is available, transfer is delayed another 6 hours. She undergoes surgery later that morning but does not regain use of her legs. She consults an attorney and a lawsuit is filed.
Plaintiff: On my first ED visit you were more focused on my chest than my main complaint of back pain (anchoring bias.) Once you saw an abnormality on my CXR, you decided you'd struck paydirt, even when the CT you ordered didn't support your initial impression (confirmation bias). I didn't have "pneumonia." On the second ED visit you assumed my earlier diagnosis was correct, did not review my hx and didn't check my sensation or reflexes, even when the nurse wrote that my sensation was only normal to my lower rib margin. This was not an azithromycin reaction (more anchoring bias). On both visits no one took a decent hx. No one did a thorough neuro exam or a rectal exam. No one had me walk. No one considered anything other than my chest complaints. When I returned by ambulance, you weren't even considering an SEA and did a risky LP before ordering an MRI. Everything took way too long. You accepted extended delays and had no sense of urgency. I was paralyzed and needed emergent surgery. Your loyalty to your hospital system's finances trumped my need for emergent surgery. You should have sent me to the nearby teaching hospital the minute you stumbled on my diagnosis and no bed was available in your own system. I would be able to walk today if not for everyone's negligence and delays.
Defense: The course of your paralysis was so fast that a secondary spinal cord infarct is by far the most likely cause. Earlier surgery would not have helped. We considered a primary neurological problem on your second visit and ruled it out. You were asking to go home. The delay of the MRI and transfer made no difference. Most studies show that outcomes are only worse if surgery is done beyond 24 hours.
Result: After expert review supporting negligence and causation, the attorney transferred the case to a law firm in another state. The outcome of the case is unknown. While the defense argument has weight, earlier diagnosis is always better and was possible in this case.
Takeaways:
1. Most often the first presentation gets a “pass.”
2. The second visit suggests that a neurological problem was considered but poorly evaluated.
3. Everyone agrees that “earlier surgery is better,” but in practice "early" means immediate.
4. SEA’s rarely move this fast, so an untreatable secondary infarct is a very likely cause. A spinal cord infarct could have played a role, but the diagnosis should have been made earlier. If so, the infarct would not have happened; earlier surgery would have led to a better outcome .
5. Think SEA in all back pain patients. It can be ruled out on H & P alone, but consideration should be documented.
6. An ESR or CRP will almost always rule in or out an SEA. More importantly, the test shows that you considered it.
7. When you suspect an SEA, don’t do an LP until after a negative MRI. An LP done through an abscess can move the infection INTRA-dural.
8. WALK YOUR PATIENTS and document it. It's the best single neuro test one can do.
Reference: Errors in Diagnosis of Spinal Epidural Abscesses in the Era of Electronic Health Records. Bhise V, Meyer A, Singh H, Wei L. Amer J Med 130(8), March 2017. Abstract and free full-text download available at link.