Spellman (2024): MRI interpretation and cauda equina compression

Spellman v Portsmouth Hospitals University NHS Trust

[2024] EWHC 2011 (KB)
High Court, King’s Bench Division
Mr Justice Sweeting
Judgment delivered: 31 July 2024

Background

The claimant had a long history of recurrent back pain, urinary and bowel symptoms and three previous lumbar operations at L4–5. Earlier investigations had not demonstrated cauda equina compression, although she had previously required intermittent self-catheterisation and had experienced sensory and continence symptoms.

On 6 June 2017, she presented with acute lower-back pain, left-leg symptoms, altered perineal sensation and abnormal sensation when passing urine. The referral information raised concern about possible cauda equina syndrome.

An urgent lumbar MRI was obtained. Image quality was reduced by significant movement artefact because the claimant was in pain. The reporting consultant radiologist nevertheless considered the examination diagnostic and reported:

multilevel lumbar disc degeneration.
a slightly tight canal at L2–3 and L3–4.
reduced cerebrospinal fluid within the thecal sac at those levels.
epidural lipomatosis contributing to thecal sac constriction.
mild central canal stenosis.
postoperative change at L4–5.

No acute cauda equina compression or surgically remediable lesion was reported. Following clinical review, the claimant was discharged.

Six days later, a private spinal surgeon reviewed the same MRI, suspected cauda equina compression and performed urgent L3–4 decompression and discectomy. The claimant subsequently continued to experience urinary and bowel dysfunction, neuropathic symptoms and back pain.

The claim

The claimant alleged that the reporting radiologist had:

understated the degree of stenosis at L2–3 and L3–4.
wrongly described the stenosis as mild rather than moderate, moderately severe or severe.
failed to identify substantial cauda equina compression.
failed to identify pathology capable of accounting for the claimant’s symptoms.
failed to recommend repeat imaging or specialist discussion.
produced an unduly reassuring report that led to delayed decompression.

The claimant contended that correct reporting would have led to surgery by the morning of 7 June 2017 and a materially better neurological and functional outcome.

The defendant denied that the MRI demonstrated cauda equina compression. It maintained that the report was reasonable and that earlier surgery would not have improved the claimant’s outcome.

Damages had been agreed at £300,000, subject to the claimant establishing both breach of duty and causation.

The central issues

The court had to determine:

what the MRI of 6 June 2017 objectively demonstrated.
whether the cauda equina was compressed.
whether the report understated or omitted material pathology.
whether the degraded examination should have been repeated.
whether the claimant had cauda equina syndrome attributable to radiological compression.
whether earlier decompression would have occurred following a different report.
whether earlier surgery would probably have improved the outcome.

The claimant accepted that if the MRI did not demonstrate cauda equina compression, failure to report compression could not amount to negligence and no injury could have resulted from that alleged omission.

The legal approach

The court applied the ordinary principles in Bolam v Friern Hospital Management Committee, subject to the logical scrutiny required by Bolitho v City and Hackney Health Authority.

Following Penney v East Kent Health Authority and Brady v Southend University Hospital NHS Foundation Trust, the court adopted a two-stage analysis:

What did the images show?
This was a question of fact for the court, decided on the balance of probabilities with the assistance of expert evidence.
Was the report reasonably competent?
Once the imaging findings had been determined, the court had to assess whether the interpretation and communication fell within a responsible and logically defensible body of radiological practice.

A court may therefore find that imaging contained a particular feature without necessarily finding that failure to report it was negligent. Conversely, conflicting expert opinions do not require the court to accept that both interpretations were professionally defensible.

What the MRI showed

The court found that the MRI demonstrated:

moderate central canal stenosis at L2–3 and L3–4.
facet and ligamentous hypertrophic degeneration.
epidural lipomatosis contributing to thecal sac constriction.
a left intracanalicular facet-joint cyst at L2–3, although this was not distinctly demonstrated prospectively.
a small left paracentral disc extrusion at L3–4.
reduced, but not absent, cerebrospinal fluid signal at the affected levels.
longstanding constitutional and degenerative change.

Crucially, the court found no point at which the nerve roots were sufficiently crowded or clumped to demonstrate compression of the cauda equina.

The presence of stenosis, epidural fat, a facet-joint cyst or a small disc abnormality did not itself prove neural compression. The decisive question was their effect on the cauda equina, not merely whether they occupied space within an already narrowed canal.

Constriction was not compression

The judgment distinguished between:

narrowing of the spinal canal.
constriction or deformation of the thecal sac.
actual compression of the cauda equina nerve roots.

Although the thecal sac was narrowed at L2–3 and L3–4, cerebrospinal fluid remained visible around the nerve roots. The court accepted the defendant’s expert evidence that the nerve roots were not clumped into a compressed mass.

Accordingly, the anatomical abnormalities did not provide radiological evidence of acute cauda equina compression or a surgically correctable cause for the claimant’s presentation.

Mild or moderate stenosis

The expert radiologists agreed that grading central canal stenosis is substantially subjective. Terms such as “mild”, “moderate” and “severe” are not governed by a universally accepted quantitative threshold.

Published grading systems differed in:

the anatomical features assessed.
whether measurements or morphology were emphasised.
the terminology assigned to particular appearances.
the relationship assumed between narrowing and clinical symptoms.

The court accepted that “moderate” was a better description of the stenosis than “mild”. However, the distinction did not establish negligence.

The original report also recorded:

that the canal was tight.
that less cerebrospinal fluid was visible.
that epidural fat materially contributed to constriction.
that there was multilevel degenerative disease.

Read as a whole, the report did not materially misrepresent the imaging. In the absence of a fixed grading standard, describing the stenosis as mild rather than moderate remained within the range reasonably open to a competent radiologist.

More importantly, both spinal experts agreed that moderate stenosis would ordinarily be managed conservatively rather than by emergency decompression. The terminology therefore did not establish a causative reporting error.

Image quality and the decision not to repeat the MRI

The MRI was degraded by substantial movement artefact. However, poor image quality did not automatically make the examination non-diagnostic.

The reporting radiologist:

recognised the artefact.
understood that the examination had been requested urgently to assess possible cauda equina syndrome.
began reviewing the images while scanning continued.
asked for an important sequence to be repeated.
considered whether further imaging would add useful information.
concluded that additional attempts were unlikely to produce materially better images while the claimant remained in significant pain.

Both radiology experts agreed that a responsible and reasonably competent body of radiologists could regard the examination as sufficiently diagnostic to issue a definitive report without recommending another scan.

The court therefore found that:

it was reasonable to interpret and report the available study.
there was no duty to continue scanning indefinitely.
further imaging offered uncertain benefit and additional discomfort.
failure to recommend repeat MRI or external specialist review was not negligent.

The claimant’s radiology expert

The claimant’s expert originally identified a significant disc extrusion at L3–4 as the cause of severe cauda equina compression. He later accepted that he had:

identified pathology at the wrong spinal level.
mistaken a facet-joint cyst at L2–3 for a disc extrusion at L3–4.
failed initially to identify the separate small L3–4 disc abnormality later relied upon.
overstated the stenosis as severe before agreeing that moderate was appropriate.
changed his opinion about whether the small disc abnormality required reporting.
accepted that the facet-joint cyst could reasonably have been missed prospectively.
withdrawn the contention that repeat imaging or discussion with a regional spinal centre was required.

He had reviewed converted JPEG images rather than the original medical-grade images on radiological software. The defendant’s expert had been provided with the same material but did not make comparable anatomical or localisation errors.

The court found that these errors reduced confidence in the claimant expert’s interpretation. Where the radiologists disagreed about the images, it preferred the evidence of the defendant’s expert.

The defendant’s radiology expert

The defendant’s expert considered the abnormalities longstanding, constitutional and degenerative. In his opinion:

cerebrospinal fluid remained present at the affected levels.
the narrowing did not produce sufficient crowding to compress the cauda equina.
the facet-joint cyst formed part of the overall moderate stenosis.
the small disc abnormality was not compressive.
the imaging appearances were commonly encountered without cauda equina syndrome.
there was no acute, surgically remediable lesion.

The court found his anatomical explanation logical and consistent with the images. It accepted that stenosis and space-occupying pathology do not inevitably interfere with nerve function while sufficient cerebrospinal fluid and space remain around the nerve roots.

The spinal-surgical evidence

The claimant’s spinal expert considered that an apparent absence of cerebrospinal fluid on one image indicated compression and that a repeat scan or urgent surgery was required. However, this produced a tension in his evidence: the scan was said to be too poor for reliable surgical planning but sufficiently diagnostic to mandate urgent decompression.

The defendant’s spinal expert considered that:

the scan was difficult to interpret but did not demonstrate cauda equina compression.
an isolated reduction in cerebrospinal fluid signal could reflect artefact or postoperative fibrosis.
genuine compression would be expected to affect adjacent images more progressively.
the claimant had previously experienced similar symptoms without radiological compression.
pain, distress and pre-existing dysfunction provided alternative explanations for the presentation.

Both spinal experts agreed that moderate stenosis alone would usually lead to conservative outpatient management rather than emergency intervention.

Clinical symptoms and imaging

Cauda equina syndrome is a clinical diagnosis assessed through symptoms, examination and imaging. Urgent MRI remains necessary where red-flag symptoms are present, but such symptoms do not by themselves prove compressive cauda equina syndrome.

The court noted that:

many patients following a suspected cauda equina pathway have no explanatory compression on imaging.
clinical symptoms correlate imperfectly with the severity of lumbar stenosis.
the claimant had experienced similar bladder, sensory and pain symptoms previously without cauda equina compression.
her prior urinary and bowel history was complex.
the history given in June 2017 contained material inconsistencies.
the alleged fall was not disclosed to the original treating clinicians.
a later episode in 2018 produced similar symptoms without explanatory imaging.
there was no reliable evidence of material neurological progression between 6 and 12 June 2017.

The symptoms therefore did not provide a sufficient basis for inferring compression that was not demonstrated radiologically.

The later operation

The private surgeon’s decision to operate on 12 June 2017 did not establish that the MRI had demonstrated acute cauda equina compression six days earlier.

The court noted that:

the surgeon did not give evidence.
no detailed contemporaneous history had been recorded.
his correspondence and operative documentation differed regarding episodes of urinary incontinence.
it was unclear precisely what pathology he considered diagnostic.
decompression was performed only at L3–4, despite alleged abnormalities at both L2–3 and L3–4.
no repeat preoperative MRI was obtained.
the claimant’s symptoms persisted after surgery.

The court made no criticism of the private surgeon’s professional judgment. It held only that the later decision to operate did not prove the factual or legal propositions required in the claim.

Breach of duty

The court found no breach of duty.

The MRI contained moderate stenosis and additional degenerative abnormalities, but it did not demonstrate cauda equina compression. As compression was not present radiologically, failure to report it was an accurate assessment rather than a negligent omission.

The report was also reasonable in its wider description because:

the scan was diagnostically adequate despite movement artefact.
reduced cerebrospinal fluid was recorded.
epidural fat and its contribution to constriction were identified.
postoperative and degenerative changes were described.
the distinction between mild and moderate stenosis was subjective.
the facet-joint cyst was not sufficiently distinct to require prospective identification.
the small disc abnormality was not compressive or clinically material.
neither repeat imaging nor external consultation was required.

Causation

Because breach of duty was not established, the causation claim necessarily failed.

The evidence also created substantial independent difficulties for causation:

moderate stenosis would not ordinarily have prompted emergency surgery.
no compressive surgical target was shown.
the claimant’s symptoms had several plausible alternative explanations.
the timing and progression of symptoms were unclear.
the claimant had similar previous presentations without compression.
the later operation did not produce the alleged improvement.
some symptoms may have pre-dated the index presentation.
the evidence did not establish that decompression on 7 June would probably have produced a better outcome.

The claim was dismissed.

Significance for radiological practice

Identify the clinical question. In suspected cauda equina syndrome, the central radiological task is to identify or exclude compression and any surgically remediable cause.
Distinguish narrowing from neural compression. Stenosis, thecal sac constriction and reduced cerebrospinal fluid do not necessarily mean that the cauda equina is compressed.
Describe the functional effect of pathology. The report should explain whether abnormalities crowd, displace or compress the nerve roots.
Read grading terms in context. “Mild”, “moderate” and “severe” may lack fixed boundaries and should not be considered in isolation from the report’s anatomical description.
Do not overstate isolated findings. A cyst, disc protrusion or epidural fat may be present without constituting an acute surgical target.
Assess the whole examination. A single degraded or apparently alarming image should be correlated with adjacent levels, sequences and planes.
Record technical limitations. Movement artefact should be acknowledged, together with its effect on diagnostic confidence.
Repeat imaging only where necessary. A technically imperfect scan may still answer the urgent clinical question.
Avoid retrospective reconstruction. Later imaging and surgery may assist interpretation but cannot automatically establish what should have been reported prospectively.
Communicate material uncertainty. If image quality prevents confident exclusion of compression, that limitation and any required next step should be stated expressly.

Significance for expert witnesses

Review the original medical-grade images wherever possible.
Confirm spinal levels systematically before attributing pathology.
Distinguish a facet-joint cyst, disc abnormality, epidural fat and postoperative change.
Separate the presence of stenosis from its effect on the cauda equina.
Explain whether cerebrospinal fluid is reduced or absent.
Avoid inferring preoperative findings uncritically from postoperative imaging.
State whether grading terminology reflects an accepted system or subjective clinical usage.
Reconcile any contention that imaging is non-diagnostic with reliance on the same imaging to mandate surgery.
Identify clearly when an opinion has changed and why.
Address whether any alleged reporting difference would probably have altered clinical management.
Keep breach and causation analytically separate.
Test the opinion against plausible alternative explanations for the clinical presentation.

Practical framework for reviewing a suspected cauda equina MRI

Was the clinical indication and urgency understood?
Was the examination technically adequate to answer the clinical question?
What levels demonstrate canal stenosis?
What structures are causing the narrowing?
Is cerebrospinal fluid preserved, reduced or absent?
Are the nerve roots separate and freely distributed, or crowded and clumped?
Is there actual cauda equina compression?
Is there an acute or surgically remediable lesion?
Are the abnormalities acute, chronic or indeterminate?
Do findings persist across adjacent slices, sequences and planes?
Could an apparent abnormality be explained by movement or other artefact?
Would repeat imaging probably provide material additional information?
Does the report communicate the degree, cause and effect of the stenosis?
Would alternative terminology materially change clinical management?
Is any retrospective interpretation being influenced by later imaging, surgery or outcome?

Educational case commentary only. It is not legal or medical advice and is not a substitute for the full judgment. The publication date above is the date of this commentary, judgment was delivered on 31 July 2024.