Direct vs circumstantial evidence
In criminal law, evidence is classified as either direct or circumstantial. Direct evidence proves a fact without needing an inference — an eyewitness account of the crime, a confession, CCTV of the act. Circumstantial evidence requires the jury to infer the fact from other established facts — opportunity, motive, pattern, post-event conduct.
Circumstantial evidence is not inferior evidence. Juries can convict on circumstantial evidence alone, and do so regularly. But the UK courts have long recognised that circumstantial-evidence cases require specific judicial direction: the jury must consider whether the circumstances, taken together, are consistent only with guilt and inconsistent with any other reasonable explanation.
The Pollock test
The canonical English formulation is R v Exall (1866) per Pollock CB: in a circumstantial case the prosecution must prove the guilt of the defendant so that no other reasonable explanation is possible. The “one strand” formulation (a circumstantial case is like a rope of many strands, strong because of the number of strands) comes from the same line of authority.
Modern formulations from the Court of Appeal require the jury to: (1) consider each circumstance individually; (2) consider the circumstances collectively; (3) ask whether, taken together, they are consistent only with guilt; and (4) be satisfied beyond reasonable doubt that no other reasonable explanation fits.
How the Letby case fits the framework
The Letby case is circumstantial: no eyewitness direct account of any act; no confession in police interview; no retained physical exhibits; no forensic-standard toxicology or pathology. The Crown’s case rests on inference from:
- Dr Evans’s causation opinion on clinical signs.
- The shift-rota overlap pattern.
- The Roche Cobas insulin screening result.
- The Post-it notes and Facebook searches.
- The handover sheets retained at home.
Each strand is itself an inference, not a direct finding. The Crown’s framing is that the strands, taken together, are consistent only with guilt.
Where the framework fails on the current evidence
The Pollock test requires the jury to be satisfied that the circumstances are consistent only with guilt and inconsistent with any other reasonable explanation. The post-conviction record establishes that an alternative reasonable explanation exists for each strand:
- Causation. The Shoo Lee Panel provides reasonable alternative natural-cause explanations for every indicted case.
- Shift-rota overlap. Selection-effect analysis (Gill, Green, Hutton, Fenton, Schneps) provides reasonable alternative explanation for the pattern.
- Insulin assay. The Joint Insulin Report provides reasonable alternative explanations including auto-antibodies, sepsis, assay interference.
- Post-it notes. Clinical-psychology expert reports provide reasonable alternative explanation: self-blame under sustained accusation.
- Facebook searches. Base-rate analysis provides reasonable alternative explanation: normal nursing behaviour.
- Handover sheets. NHS confidentiality framework provides reasonable alternative explanation: professional practice.
On the current evidence, the strands are not “consistent only with guilt”. Each has a reasonable alternative explanation backed by reputable expert opinion. The Pollock test is therefore not satisfied on the current evidence.
The Cannings principle as circumstantial-evidence corollary
The Cannings principle (Court of Appeal, December 2003) is, in effect, the specific application of the Pollock test to medical-evidence cases: where circumstantial medical evidence is disputed by reputable experts, the inference does not reach “consistent only with guilt”. The Letby case satisfies the Cannings condition in its strongest form.
Why this matters for appellate review
The Court of Appeal, on referral, will apply the Pollock test to the evidence now before it — not the evidence of 2023. On the current evidence, the test is not satisfied. This is the legal-framework restatement of the scientific-framework conclusion the Shoo Lee Panel reached.
What circumstantial evidence can carry
UK criminal-evidence doctrine recognises circumstantial evidence as capable of carrying conviction at the criminal standard (beyond reasonable doubt) where the cumulative pattern excludes any alternative reasonable explanation. The classical statement is that circumstantial evidence is like a rope: each strand may be weak, but woven together the rope can be strong enough to bear significant weight. The doctrine has historically been applied across many categories of UK criminal cases.
The doctrinal limit is that the cumulative pattern must actuallyexclude alternative reasonable explanations. Where alternative reasonable explanations are not excluded by the evidence, the rope-of-strands metaphor fails: each strand is weak, and the cumulative weakness does not produce strength. The Letby case’s circumstantial framework engages this doctrinal limit directly.
The Letby circumstantial structure
The Crown’s circumstantial case at trial included: the shift-rota chart; the handover-sheet retention; the Facebook-search pattern; the handwritten notes; the Crown’s ‘preparation evidence’ framework. Each element was presented as part of a cumulative pattern. The post-conviction critique addresses each element on its own terms (the denominator analysis on Facebook searches; the 257/21/236 ratio on handover sheets; the self-blame psychology framework on the notes; the selection-bias framework on the shift chart) and concludes that the cumulative pattern fails on its own internal logic when the elements are evaluated against documented baselines.
What the Cannings principle adds
The Cannings principle (2004) added a specific overlay to circumstantial evidence in medical-causation cases: where serious disagreement exists between reputable experts on the medical causation of death, circumstantial evidence cannot displace that disagreement to ground a criminal conviction. The Letby case’s circumstantial structure proceeded as if the Crown’s medical-causation experts were institutionally authoritative. The Panel finding establishes that they were not. The circumstantial structure therefore engages both the classical-circumstantial-doctrine limits and the Cannings-principle overlay.
Why this framework matters for the CCRC review
A judicial weighing of the Letby case under the ‘real possibility’ test will engage both the medical-causation evidence and the circumstantial-evidence framework. The CCRC review will weigh whether the circumstantial pattern, evaluated against the documented baselines and the serious-expert-disagreement on the medical-causation evidence, can still carry conviction at the criminal standard. The structural argument is that it cannot.