Expert “Medical Torture” Diagnoses Are Inadmissible in Torture Trials When They Are Unhelpful, Unreliable, and Function as a Legal Conclusion

I. Introduction

In People of Michigan v. Gwendolyn Josephine Alexander (Mich Sup Ct, July 15, 2026), the Michigan Supreme Court confronted an evidentiary problem that increasingly arises in criminal trials that feature specialized medical testimony: whether clinically phrased diagnostic labels can nonetheless operate as de facto legal conclusions.

The defendant was convicted of, among other offenses, torture under MCL 750.85 after evidence showed her six-year-old son (MA) suffered ligature marks, bruising, swelling, and scarring, and after defendant admitted MA had been restrained for hours at a time. At trial, a child-abuse pediatric expert, Dr. Dena Nazer, testified she diagnosed MA with “medical torture”—a clinical label she described as requiring at least two physical assaults accompanied by at least two forms of psychological maltreatment.

On leave-related argument, the Supreme Court focused on a narrow but consequential question: whether the expert’s use of the phrase “medical torture” invaded the province of the jury in a case where the charged crime was torture. Although the Court held admission of the diagnosis was erroneous under Michigan’s expert evidence rules, it ultimately affirmed because the error was unpreserved and did not satisfy the prejudice requirement of plain-error review.

II. Summary of the Opinion

  • Error: Dr. Nazer’s “medical torture” diagnosis testimony was inadmissible under MRE 702 because it was unhelpful to deciding a fact in issue, lacked a reliable foundation (especially as to psychological maltreatment), and invaded the province of the jury by coming too close to a legal conclusion on the charged offense of torture.
  • No relief: Because the defense did not object, the Court applied plain-error review and held defendant failed to show the error affected the outcome given the strength of properly admitted evidence (medical observations, child testimony, and defendant’s admissions).
  • Key doctrinal move: The Court extended the logic of People v McFarlane—which disapproved “abusive head trauma”/“child abuse” diagnostic labels when they convey willfulness and culpability—to “medical torture” in a torture prosecution.

III. Analysis

A. Precedents Cited (and How They Shaped the Decision)

1. Framing the standard of review and the evidentiary lens

  • People v Thorpe: The Court relied on Thorpe for two central principles: (i) expert evidence is powerful and can mislead because jurors give it special weight, and (ii) experts may not vouch for credibility. Thorpe supplied the warning that juries look to experts as an “objective source” and a “hook,” amplifying the danger of conclusory diagnostic labels.
  • People v McDaniel: Used for the proposition that preliminary legal questions embedded in evidentiary rulings are reviewed de novo (though the overall admission decision is discretionary).
  • People v Bynum: Established that because there was no trial objection to the “medical torture” phrasing, the issue was unpreserved and review was for plain error.

2. The “helpfulness” and “reliability” gatekeeping requirements

  • People v Kowalski: Provided the Court’s central analytic structure under MRE 702: “helpfulness” is a threshold inquiry, and unreliable opinions are unhelpful. Kowalski underwrote the Court’s conclusion that a diagnosis can be excluded when it does not assist the jury on a disputed, technical matter.
  • People v Smith and People v Thorpe: Together supplied the reliability rule that an examining physician cannot give an opinion that is effectively just an endorsement that the complainant is telling the truth, absent expertise in assessing credibility. The Court applied Smith/Thorpe to the “psychological maltreatment” component of “medical torture,” finding it rested largely on crediting MA’s narrative and “other information,” rather than an independently reliable medical foundation.

3. Ultimate-issue testimony versus legal conclusions and jury province

  • Downie v Kent Prod, Inc, MRE 704, and People v Smith: The Court reaffirmed that testimony may embrace an ultimate issue, but only if otherwise admissible—meaning it must still satisfy all evidentiary constraints.
  • People v Robinson: Cited for the core prohibition on experts opining on guilt or innocence.
  • People v Young: Cited for the “province of the jury” concept: experts may not usurp the jury’s role by effectively telling it how to decide central issues of culpability.
  • People v Drossart, People v Parks, and People v Fomby: Reinforced the maxim that experts may not testify about legal requirements as applied to the facts, nor phrase opinions as legal conclusions, and may not express opinions tantamount to guilt.

4. The controlling analogy: “abuse” labels in abuse prosecutions

  • People v McFarlane: The Court treated McFarlane as the closest doctrinal analog. McFarlane permitted experts to explain that injuries were “inflicted” rather than accidental, but condemned diagnostic labels like “abusive head trauma” and agreement that injuries amounted to “child abuse,” because “abuse” implies willfulness and moral culpability and thus implicates intent/criminal responsibility. The Alexander Court held “torture,” like “abuse,” carries similar lay connotations of purposeful cruelty.
  • People v McFarlane, 505 Mich 1059 (2020) (CAVANAGH, J., concurring): Quoted to emphasize the specific danger that medical terminology can confuse a medical diagnosis with a legal determination.

5. Prejudice and plain-error relief

  • People v Carines: Supplied the three-part plain-error test (error, clear/obvious, outcome effect) and the final discretionary reversal standard (fairness/integrity/public reputation).
  • People v Randolph: Used to explain why an error may not be “clear or obvious” if subject to reasonable dispute—here, because McFarlane involved different terminology and charges.
  • People v Davis: Supported the presumption that jurors follow instructions (relevant to the Court’s lack-of-prejudice conclusion).

6. The prejudice/probative balance backdrop

  • People v Peterson and People v Beckley: These cases supplied the Court’s broader caution: expert evidence is persuasive, and as reliability diminishes, prejudice increases. Although the Court did not center its analysis on MRE 403, it used Peterson/Beckley to explain why emotionally charged diagnostic labels are uniquely risky.
  • People v Sharpe: Cited to illustrate “otherwise admissible” analysis across the rules of evidence—supporting the Court’s insistence that expert evidence must satisfy multiple evidentiary checks.

B. Legal Reasoning

1. MRE 702 “helpfulness”: why the “medical torture” label did not assist the jury

The Court accepted that medical experts can be necessary to help jurors understand the nature, extent, timing, and mechanism of injuries—especially as to whether “great bodily injury” occurred, an element of torture under MCL 750.85(1). But it drew a sharp line between (i) medically descriptive testimony that translates physical findings for lay jurors and (ii) a high-level diagnostic label that does not advance a legally relevant factual inquiry.

Dr. Nazer’s “medical torture” label did not explain the injuries’ timing, severity, or mechanism in a way that helped decide “great bodily injury.” By contrast, testimony describing scars as linear, hyperpigmented, circumferential, and consistent with ligatures/zip ties did help. This contrast was central to the Court’s conclusion that the diagnosis itself was unhelpful and thus inadmissible.

2. Reliability: why the psychological-maltreatment prong lacked a dependable foundation

The Court found the “psychological maltreatment” component was not rooted in independently verifiable medical methodology, but largely in crediting MA’s account and “other information” from the Child Advocacy Center process. Under People v Smith and People v Thorpe, when an opinion depends on believing a complainant’s narrative, and the expert is not qualified to assess credibility, the opinion lacks a reliable foundation. Because “medical torture” required psychological maltreatment findings, the unreliability of that component further undermined the diagnosis.

3. Province of the jury: why “medical torture” functions like an opinion on legal culpability

The opinion’s most significant doctrinal move is its equation of “torture” with “abuse” for jury-province purposes. The Court reasoned that “torture” in ordinary usage connotes intentional infliction of intense pain to punish/coerce/satisfy cruelty—closely tracking the statute’s intent element: intent “to cause cruel or extreme physical or mental pain and suffering.” Thus, even if framed as “medical,” the diagnosis risked conveying to jurors that an expert had effectively identified the legal wrong at issue, encouraging deference to an authority figure on the very question the jury must decide.

This is the McFarlane principle applied beyond “abusive head trauma”: diagnostic labels that embed moral blame or willfulness become legally dangerous when they echo the charged offense. The Court also rejected the idea that the prosecutor’s limited reliance on the label in closing could cure the underlying admissibility problem; that point went to prejudice, not error.

4. Plain error: why the convictions were affirmed despite the evidentiary violation

Even assuming the admission was “plain,” the Court held there was no outcome effect because the remaining evidence was strong and largely redundant of the improper label: multiple witnesses observed ligature marks; MA and RS described zip-tying; and defendant admitted she restrained MA and directed Scott to do so. The expert also conceded she could not speak to the legal definition of torture, and the jury received standard instructions on its role in evaluating testimony. Accordingly, defendant failed the Carines prejudice requirement.

C. Impact

1. Practical rule for trial courts and litigants

Alexander establishes that Michigan courts should treat clinically used terms that mirror charged crimes—here, “medical torture”—as presumptively problematic when they: (1) do not help the jury decide a fact in issue under MRE 702, (2) rest on credibility-dependent foundations, or (3) risk functioning as a legal conclusion on intent or culpability. The decision encourages prosecutors to elicit descriptive medical testimony (mechanism, timing, severity, consistency with ligatures) rather than morally freighted labels. Defense counsel, conversely, has a clear roadmap for timely objection and for insisting on sanitized terminology.

2. Extension of McFarlane beyond “abuse” terminology

The most enduring doctrinal effect is that McFarlane’s “diagnostic-label-as-legal-conclusion” logic is not confined to “abuse” or “abusive head trauma.” Alexander signals that any medical label that strongly overlaps with a statutory offense and implies willfulness (e.g., terms that connote cruelty, punishment, coercion, sadism) may be excluded when offered as an expert “diagnosis” in the corresponding criminal trial.

3. Reliability scrutiny for psychological-maltreatment opinions

The Court’s foundation analysis also pressures litigants to separate (i) medically grounded conclusions from physical findings from (ii) psychosocial conclusions drawn mainly from interviews and collateral sources. Where an expert’s “diagnosis” requires psychological components, Alexander suggests courts must ask whether the methodology is reliably medical or merely a repackaged credibility judgment barred by Smith/Thorpe.

IV. Complex Concepts Simplified

  • MRE 702 (“helpfulness” and “reliability”): Experts may testify only if their specialized knowledge actually helps jurors decide something they cannot decide well on their own, and only if the opinion is grounded in reliable methods and data. If the opinion is essentially “I believe the child,” it is not reliable expert evidence.
  • “Invading the province of the jury”: The jury alone decides guilt and the defendant’s intent. An expert cannot indirectly answer those questions by using labels that imply the legal conclusion (e.g., “torture”).
  • Ultimate issue (MRE 704) vs. legal conclusion: Even if an opinion touches an ultimate question, it still must be admissible under other rules; experts cannot testify in a way that effectively states “the defendant committed the crime.”
  • Plain error (People v Carines): If there is no objection at trial, a defendant must show not only that the judge made an error, but also that the error likely changed the outcome. Strong independent evidence often defeats relief even when an error occurred.

V. Conclusion

People v Alexander clarifies and extends Michigan’s expert-testimony gatekeeping in criminal cases: a clinician’s diagnostic label is not admissible simply because it is “medical.” When a diagnosis is unhelpful to the elements in dispute, rests on unreliable credibility-dependent foundations, or tracks the charged offense in a way that implies intent and culpability, it crosses from expert assistance into jury usurpation—especially given the persuasive power of expert testimony.

At the same time, Alexander is a cautionary plain-error case: even significant evidentiary mistakes may not yield reversal absent a timely objection and a concrete showing that the verdict likely turned on the improper evidence.