Lit Matters 3: CT sensitivity for pyogenic spinal infections

Pyogenic spinal infection is an ED back-pain diagnosis that CT often misses, especially spinal epidural abscess. In patients already concerning enough to undergo MRI, CT was reasonably sensitive for infections outside the canal but performed poorly for SEA, making MRI the imaging test of choice when suspicion is real.

CT Versus MRI for Spinal Infection

• MRI first-line imaging: Pyogenic spinal infection remains an MRI diagnosis in emergency care; when clinical concern is meaningful, starting with CT risks false reassurance and delayed recognition of epidural disease.
• CT sensitivity gap: CT detected 79% of pyogenic spinal infections overall in masked review, but that headline number hides the main problem: performance collapsed for spinal epidural abscess.
• SEA detection failure: Spinal epidural abscess was seen on CT only 18% of the time, reinforcing that a negative CT does not come close to excluding the highest-stakes spinal canal infection.
• Outside-the-canal infections: CT was better for vertebral, disc, paravertebral, and other extra-canal disease, reaching 83% sensitivity for infections outside the spinal canal, a distinction worth hearing in the episode.
• Initial read limitations: Real-world first reads underperformed the study overread, with just 50% overall sensitivity, which matters far more at 2 a.m. than idealized retrospective image review.

Clinical Pattern of Pyogenic Spinal Infection

• Overlapping infection syndromes: Pyogenic spinal infections commonly travel together rather than appearing in isolation; finding one lesion should immediately raise concern for a second spinal infectious focus.
• SEA co-infection signal: Most spinal epidural abscess cases were accompanied by another spinal infection, with 82% of SEA patients having an associated lesion somewhere else in the spine.
• Paravertebral abscess association: Paravertebral abscess or infection tracked especially closely with epidural disease: every patient with PVA in this cohort also had a spinal epidural abscess.
• Septic facet warning: Septic facet joint infection is uncommon but not benign; more than 20% of patients with septic facet infection also had a spinal epidural abscess.
• Whole-spine implications: Noncontiguous lesions are part of the concern once one spinal infection is identified, though the practical MRI extent questions and ED tradeoffs are where we get into nuance in the episode.