Date: October 20, 2023

Reference: Cai et al. Implementation of a Clinical Management Tool for Spinal Epidural Abscess Early Diagnosis. AEM October 2023.

Guest Skeptic: Dr. Kirsty Challen is a Consultant in Emergency Medicine at Lancashire Teaching Hospitals.

Case: You are in your group meeting and have heard about a case at a nearby emergency department (ED) where the diagnosis of a spinal epidural abscess was delayed, and a substantial settlement has been made out of court. Your group director is concerned with avoiding the same thing happening in your department and wants to know if you should implement an evidence-based clinical management tool (CMT) to reduce delays in diagnosis.

Background: Spinal epidural abscess (SEA) is a diagnosis which can seem easy to make in retrospect. The majority of time (55%) the diagnosis of  SEA often involves an error with a median length of time to diagnosis of 12 days according to one study [1].  Diagnostic delays were found to be present in 75% of SEA patients with only a minority (10-15%) of patients present with the “classic triad” of fever, back pain and neurologic deficit [2].  Another study reported that the 90% of patients misdiagnosed on their first ED visit [3] likely due to the non-specific and variable initial presentation, and the number of patients with back pain of benign origin seen in EDs [4].

SEA is the condition with proportionately the highest misdiagnosis rate in ED per a recent systematic review, and long-term sequelae for patients with associated medico-legal costs are high [5]. However, there is a need for clinicians to not let SEA become the next pulmonary embolism with high rates of over investigation.

We’ve looked at back pain on the SGEM before, mostly in terms of treatments. SGEM#366 concluded that we could not recommend the routine use of skeletal muscle relaxants, SGEM#304 agreed that adding acetaminophen to ibuprofen did not improve one-week outcomes, while SGEM#173 concluded the same about diazepam (there’s a theme here!).

We’ve also looked at the rational use of imaging on SGEM#283 (the Ottawa subarachnoid haemorrhage rule is highly sensitive but has very poor specificity). In SGEM#181, we were unconvinced of the value of routine use of whole body CT in trauma patients, and right back in 2015 on SGEM#106 discussing the Canadian CT head rule and the New Orleans Criteria. However, we’ve not previously looked at the intersection of non-traumatic back pain and rational investigation.


Reference: Cai et al. Implementation of a Clinical Management Tool for Spinal Epidural Abscess Early Diagnosis. AEM October 2023.

  • Population: Adults attending at continuously-staffed EDs in a health network covering 15 states, 2016-19.
    • Excluded: Facilities not collecting radiology or lab order data, facilities without 6 months of data before and after intervention.
  • Intervention: Implementation of a literature-based Clinical Management Tool (CMT).
  • Comparison: Periods before and during the implementation of the CMT.
  • Outcomes:
    • Primary Outcome(s): 1. The proportion of patients with SEA with a potentially related visit in the previous 30 days and 2. For those with a prior visit, number of days from first visit to diagnosis.
    • Secondary Outcomes: Utilization rates for CT, MRI, Xray, ESR and CRP.
  • Type of Study: Implementation study.

Dr. Anglea Cai

This is an SGEM HOP and we are pleased to have the lead author on the show, Dr. Angela Cai. She is a Clinical Assistant Professor of Emergency Medicine at the University of Pennsylvania. She completed this work during her Innovation Fellowship at US Acute Care Solutions. Prior to that she trained at Kings County EM Residency in Brooklyn and the University of North Carolina Chapel Hill for her medical and business degrees.

Authors’ Conclusions: Back pain CMT implementation was associated with an increased rate of recommended imaging and laboratory testing in back pain. There was no associated reduction in the proportion of SEA cases with a related prior visit or time to SEA diagnosis.”

Quality Checklist for Observational Cohort Studies:

  1. Did the study address a clearly focused issue? Yes
  2. Did the authors use an appropriate method to answer their question? Yes
  3. Was the cohort recruited in an acceptable way? Yes
  4. Was the exposure accurately measured to minimize bias? Unsure
  5. Was the outcome accurately measured to minimize bias? Unsure
  6. Have the authors identified all-important confounding factors? No
  7. Was the follow up of subjects complete enough? Yes
  8. How precise are the results? Confidence intervals are relatively broad, from minus 4.5 to plus 6% for the change in primary outcome.
  9. Do you believe the results? Yes
  10. Can the results be applied to the local population? Yes/No
  11. Do the results of this study fit with other available evidence? No
  12. Funding? No funding source reported, although one author reports payments from 3 companies for unrelated work.

Results: There were 8.3 million ED visits in 59 EDs over the three years. One in twenty, or 333,517 of those visits were for back pain. The percent of back pain visits stayed about the same before and after the CMT was introduced (4.8% to 4.5%). However, the percent of the total ED patients diagnosed with SEAs increased from 0.006% to 0.009% after the introduction of the CMT. The percentage of the back patients diagnosed with SEA also increased from 0.13% to 0.19%


  • Primary Outcome(s):
    • Proportion of patients with SEA with a potentially related visit in the previous 30 days 12.2% vs 13.3%, difference +1.0% (95% CI; -4.5 to 6.5)
    • For those with a prior visit, number of days from first visit to diagnosis 15.2 days vs 11.9 days, difference −3.3 days (95% CI; −7.1 to 0.6)
  • Secondary Outcomes: Rates of CT, MRI, ESR and CRP testing rose, while spine Xray’s fell.

Listen to the SGEM podcast to hear Angela respond to our five nerdy questions.

1) The Clinical Management Tool (CMT) – Could you talk us through how the clinical management tool (CMT) you used risk stratifies patients?

2) Study Design – You performed a retrospective implementation study rather than an experimental design such as a cluster randomized trial. Could you talk us through your reasons for this?

3) Confounding Factors – You explored changes in gender and age distribution across the implementation periods, but you haven’t presented data on other potentially relevant factors, like injecting drug use or pre-presentation opioid prescription. Would you have liked to explore these factors too and do you think they might have affected your outcomes?

4) Potentially Related Diagnoses – You developed a consensus list of potentially related diagnoses for patients who had had an ED attendance in the prior 30 days. I was interested to see that none of them related to non-specific viral myalgia type syndromes. Was this a deliberate choice?

5) The “So What” Question – You found increased use of CMT-recommended testing (both cross-sectional imaging and blood work) but no change in delayed diagnosis. So, the “million” dollar question is – did you need a CMT?

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We agree with the authors’ conclusions.


Case Resolution: You suggest to your group director that the evidence in this field is mixed and before you change policy you should look at your local data.

Dr. Kirsty Challen

Clinical Application: There is no national recommendation for the assessment of potential SEA in the UK so I will be taking this information back to my home institution to look at how we manage it locally.

What Do You Tell the Director: There isn’t a straightforward solution to this issue. The most recent evidence suggests that a CMT changes processes but not outcomes. I think we should find out what our current local practice is first.

Keener Kontest: Last weeks’ winner was Physician Assistant Dave Michaelson. He knew Rudolf Virchow was expelled from the Charité Hospital for his political activism and liberal views, particularly his involvement with the German Revolution of 1948.

Listen to the SGEM podcast for this weeks’ question. If you know, then send an email to with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

SGEMHOP: Now it is your turn SGEMers. What do you think of this episode on a clinical management tool for spinal epidural abscess? Tweet your comments using #SGEMHOP.  What questions do you have for Angela and her team Ask them on the SGEM blog? The best social media feedback will be published in AEM.



  1. Bhise V, Meyer AND, Singh H, Wei L, Russo E, Al-Mutairi A, Murphy DR. Errors in Diagnosis of Spinal Epidural Abscesses in the Era of Electronic Health Records. Am J Med. 2017 Aug;130(8):975-981. doi: 10.1016/j.amjmed.2017.03.009. Epub 2017 Mar 31. PMID: 28366427.
  2. Davis DP, Wold RM, Patel RJ, Tran AJ, Tokhi RN, Chan TC, Vilke GM. The clinical presentation and impact of diagnostic delays on emergency department patients with spinal epidural abscess. J Emerg Med. 2004 Apr;26(3):285-91. doi: 10.1016/j.jemermed.2003.11.013. PMID: 15028325.
  3. Long B et al. High risk and low prevalence diseases: spinal epidural abscess. Am J EM 2022;53:168.
  4. Galliker G et al. Low back pain the emergency department: prevalence of serious spinal pathologies and diagnostic accuracy of red flags. Am J Med 2020;133:60.
  5. Newman-Toker et al. Diagnostic errors in the emergency department: a systematic review. AHRQ report 22:EHC043.