1) Introduction
R v Babu [2025] EWCA Crim 1870 concerned a GP (the appellant) convicted in the Crown Court at Portsmouth of
four counts of sexual assault contrary to section 3 of the Sexual Offences Act 2003, involving three patients
(identified as C1, C2, and C3). The appeal principally challenged:
- the admission of hearsay evidence about C1 (who died before trial);
- the admission of other-acts (“bad character”) evidence from five additional patients describing inappropriate but non-charged conduct;
- whether a post-conviction autism diagnosis (fresh evidence) should have altered the admissibility assessment and/or the safety of the convictions.
The Court of Appeal (Criminal Division) (Lady Justice May, Mr Justice Choudhury, and HHJ Tayton KC) refused renewed leave to appeal
against conviction and refused the application to adduce fresh evidence.
3) Analysis
A. Precedents Cited
The trial judge relied on R v Riat when assessing whether the deceased complainant’s statements were “apparently or potentially reliable”
and whether the jury could fairly evaluate reliability despite the absence of cross-examination. Riat is regularly cited for the approach to
hearsay safeguards: admission is not automatic merely because a statutory gateway is satisfied; the court must still consider reliability features and
fairness, including the jury directions needed to mitigate risks.
Here, the judge’s reasoning (endorsed by the Court of Appeal) drew directly on that safeguard-based approach: the absence of any apparent motive to lie,
no history suggesting delusion or false claims, and the ability to test reliability by reference to surrounding evidence (including the pattern of allegations).
R v Popescu was cited when considering whether the jury could retain the typed safeguarding note. The defence had argued (at trial) that the situation was
analogous to jurors retaining a transcript of an ABE interview, risking undue weight. The judge distinguished Popescu, treating this as a different category:
a contemporaneous professional record admitted so the jury could compare its contents against other evidence.
Although that ground was not pursued on renewal, the case illustrates the ongoing practical question in hearsay/document cases: whether a jury’s access to a document
enhances careful evaluation or risks distorting the evidential balance—managed through tailored directions rather than a blanket rule.
B. Legal Reasoning
i) Hearsay from a deceased complainant: statutory gateways and fairness control
The Court of Appeal’s refusal of leave effectively affirms a structured sequence that trial courts should follow:
-
Identify the gateway: C1’s statements were admitted under section 116(2)(a) CJA 2003 (death),
while the safeguarding record was admitted under section 117 (business document) and/or section 114 (interests of justice).
-
Consider reliability features: the judge found no obvious motive to fabricate and no evidence undermining C1’s capacity or truthfulness.
-
Consider how the jury can test the evidence: the judge emphasised cross-admissibility/pattern evaluation and comparison with other sources.
-
Apply exclusionary safeguards: the defence sought exclusion under section 78 PACE (and the judgment also references
section 126 CJA 2003), arguing unfairness due to inability to cross-examine and alleged flaws in the record. The judge concluded probative value outweighed prejudice,
and that careful directions could address limitations.
The Court of Appeal’s stance is notable for treating the inability to cross-examine as a serious factor but not a trump card: if the circumstances provide
sufficient indicia of reliability and the jury can evaluate the evidence in context, admission can be fair.
ii) Other-acts (“bad character”) evidence: relevance to issues and avoiding unfair prejudice
The five additional patients’ accounts were not charged as sexual assaults, but described allegedly inappropriate, overfamiliar, and unprofessional conduct.
The prosecution’s relevance case (accepted) was not merely “he is the sort of person who does this,” but that the evidence:
- contradicted parts of the appellant’s police interview stance (e.g., denials about hugging/behaviour);
- supported the complainants’ accounts by reducing the likelihood of coincidence and rebutting concoction;
- helped show an overall pattern of boundary-crossing behaviour towards patients in a clinical setting.
The fairness answer was practical: the defence could cross-examine these witnesses, allowing the jury to assess accuracy and reliability, and the trial judge would direct
the jury on permissible use. The Court of Appeal accepted the single judge’s conclusion that the evidence also bore on whether the appellant had been warned and therefore could not credibly
claim ignorance of inappropriateness (an issue made more pointed by the later autism argument).
iii) Fresh evidence of autism: materiality, timing, and the pleaded defence
The renewed application sought to introduce expert evidence (Professor Sir Simon Baron-Cohen) and evidence from the appellant’s wife to establish autism and its effect on communication.
The Court of Appeal upheld the single judge’s reasoning on two intersecting themes:
-
Could it have been obtained for trial? The court regarded the autism issue as one that could and should have been investigated earlier
(the appellant’s family circumstances and the spouse’s stated longstanding suspicions were relevant to that conclusion).
-
Would it have mattered to safety? The court treated the defence case as a decisive contextual factor: the appellant’s position was that the allegations were fabricated/false,
not that he had misunderstood boundaries due to neurodivergence. The court also noted that no communication difficulty was apparent in trial preparation or testimony and that the appellant was a highly educated professional.
On that analysis, autism evidence did not undermine the jury’s verdicts on intent/sexual nature/consent; nor did it make the admission of other-acts evidence unfair in the manner alleged.
C. Impact
Although framed as a leave decision, R v Babu is a clear marker of appellate reluctance to interfere where trial judges:
(i) carefully identify hearsay gateways; (ii) engage with R v Riat-style reliability safeguards; and (iii) manage risk through clear directions rather than exclusion.
Three practical impacts stand out:
-
Deceased-complainant cases: professional safeguarding and healthcare records may be admitted where they enable meaningful evaluation and comparison,
especially if the maker/witnesses can be cross-examined about creation and content.
-
Boundary-pattern evidence in clinician misconduct prosecutions: testimony describing non-criminal but inappropriate interactions can be admissible to rebut denial/concoction
and demonstrate a pattern relevant to the issues, provided the judge controls prejudice and avoids satellite trials.
-
Neurodivergence arguments post-conviction: a late diagnosis will not readily reopen admissibility or verdict-safety where the trial defence was outright denial and where the defendant’s functioning and prior warnings
make “social misunderstanding” an implausible alternative explanation on the facts.