Accutane Gatekeeping Applies to Differential Diagnosis in All Civil Cases: Trial Courts Must Rigorously Vet Both “Rule In” and “Rule Out” Causation Steps

1. Introduction

In Alison Beavan v. Allergan U.S.A., Inc. (Supreme Court of New Jersey, May 27, 2026), the Court addressed how New Jersey trial courts must evaluate the reliability of expert causation testimony in civil litigation—specifically where the expert relies on a differential diagnosis methodology.

Plaintiff Alison Beavan alleged that an Ozurdex intravitreal injection (from a recalled lot potentially containing a silicone particulate) caused a cascade of severe ocular events, culminating in blindness in her left eye. Defendant Allergan U.S.A., Inc. challenged the admissibility of plaintiff’s causation experts under N.J.R.E. 702 and 703 and argued their opinions were inadmissible “net opinions,” then sought summary judgment.

The trial court denied the motion to bar expert testimony and denied summary judgment without conducting the “rigorous” gatekeeping analysis prescribed in In re Accutane Litigation, 234 N.J. 340, 380-400 (2018). The Appellate Division reversed, labeling the experts’ opinions net opinions and concluding the methodology would not “pass muster” under the Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993) factors. The Supreme Court granted certification.

2. Summary of the Opinion

The Court unanimously held that Accutane requires the trial court—acting as gatekeeper—to resolve disputes over expert reliability in civil cases under N.J.R.E. 702 and 703, applying the Accutane/Daubert-derived factors if relevant. Because the trial court did not conduct that inquiry, the record was inadequate for appellate resolution; remand was required for the gatekeeping proceeding envisioned in Accutane.

  • Reliability: The case was remanded for an Accutane-compliant reliability determination (the Court “strongly encourage[d]” a N.J.R.E. 104 hearing).
  • Differential diagnosis: There is no Accutane carve-out for differential diagnosis; the gatekeeping analysis must examine both the “rule in” and “rule out” steps.
  • Net opinion: The Court reversed the Appellate Division’s ruling that plaintiff’s retained expert (Dr. Lalezary) offered a net opinion; his report sufficiently explained the “why and wherefore.”
  • Treating physician report: A treating physician designated to opine on causation must still supply a written expert report when properly requested; the trial court on remand must decide whether plaintiff may serve a late report for Dr. Phillips and, if so, whether it is admissible.
  • Summary judgment: The Court reversed the Appellate Division’s summary judgment ruling because it depended on excluding expert testimony; summary judgment may be renewed after admissibility rulings on remand.

3. Analysis

A. Precedents Cited

1) In re Accutane Litigation, 234 N.J. 340, 380-400 (2018)

Accutane is the controlling anchor of the opinion. The Court re-emphasized that New Jersey follows a methodology-based reliability inquiry and that trial courts must conduct “rigorous” gatekeeping when reliability is challenged. Critically, Accutane requires courts to assess both (i) the expert’s methodology and (ii) the underlying data used to form the opinion. The Court also reaffirmed that the Daubert factors are “perhaps pertinent for consideration, but not dispositive or exhaustive,” and New Jersey has not adopted the entire body of “Daubert jurisdiction” doctrine wholesale.

Here, the trial court did not do what Accutane demands; therefore, appellate adjudication of reliability was premature and procedurally improper.

2) Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993); General Electric Co. v. Joiner, 522 U.S. 136 (1997); Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999)

The Court used these federal authorities in the limited way Accutane permits: as sources of potentially relevant reliability factors and a conceptual framework for gatekeeping. The Appellate Division had effectively applied these considerations itself to reject plaintiff’s experts; the Supreme Court’s central correction was that such an analysis must be performed first by the trial court as gatekeeper on an appropriate record.

3) Methodology-based evolution in New Jersey: Rubanick v. Witco Chem. Corp., 125 N.J. 421 (1991) and Landrigan v. Celotex Corp., 127 N.J. 404 (1992)

These cases were cited to place Accutane (and the current decision) within New Jersey’s longstanding shift away from pure “general acceptance” toward scrutiny of methodology. They reinforce that reliability is not satisfied by credentials alone; the reasoning process and its application to data matters.

4) Rule 104 hearings and reliability procedure: Kemp ex rel. Wright v. State, 174 N.J. 412 (2002); Hisenaj v. Kuehner, 194 N.J. 6 (2008)

The Court reiterated that Accutane does not make a N.J.R.E. 104 hearing mandatory in every case, but that Kemp “solidified” the pretrial reliability hearing as a common mechanism. Hisenaj was cited for deference (abuse-of-discretion review) when the trial court has conducted “a full Rule 104 hearing,” underscoring the institutional competence and primacy of trial courts in the first-instance reliability determination.

5) Differential diagnosis in New Jersey: Creanga v. Jardal, 185 N.J. 345 (2005)

Plaintiff argued Creanga effectively validated differential diagnosis such that Accutane/Daubert analysis was unnecessary. The Court rejected that reading. Instead, it harmonized the doctrines: differential diagnosis is a permissible methodology, but its reliability remains subject to Accutane gatekeeping.

The Court imported Creanga’s two-step structure into the Accutane framework: the trial court must examine reliability at (1) the “rule in” stage (all plausible causes) and (2) the “rule out” stage (eliminating alternatives using scientific methods and procedures).

6) Federal differential diagnosis discussion cited in Creanga and reiterated here: Clausen v. M/V New Carissa, 339 F.3d 1049 (9th Cir. 2003); Heller v. Shaw Indus., Inc., 167 F.3d 146 (3d Cir. 1999); Claar v. Burlington N. R.R. Co., 29 F.3d 499 (9th Cir. 1994); Carlson v. Okerstrom, 675 N.W.2d 89 (Neb. 2004)

These cases supplied the operational content of a legally acceptable differential diagnosis: ruling in only causes “generally capable” of producing the condition, and ruling out alternatives with reasons grounded in scientific methods rather than subjective belief or speculation. The Court used these authorities to explain what, concretely, the trial court must scrutinize under Accutane when differential diagnosis is the asserted causation method.

7) Net opinion doctrine: Townsend v. Pierre, 221 N.J. 36 (2015); Polzo v. County of Essex, 196 N.J. 569 (2008); Buckelew v. Grossbard, 87 N.J. 512 (1981); Borough of Saddle River v. 66 E. Allendale, LLC, 216 N.J. 115 (2013); plus cited Appellate Division formulations Ehrlich v. Sorokin, 451 N.J. Super. 119 (App. Div. 2017) and Harte v. Hand, 433 N.J. Super. 457 (App. Div. 2013)

The Court separated two often-confused inquiries: (i) reliability under N.J.R.E. 702/703 (governed by Accutane) and (ii) whether the expert provided the “why and wherefore” (net opinion). It held Dr. Lalezary was not a net opinion witness because his report articulated the factual basis and inferential chain (recalled lot; timing; clinical course), even though specific causation and alternative causation issues remain to be vetted under Accutane on remand.

8) Civil procedure sequencing and review standards: Townsend v. Pierre, 221 N.J. 36 (2015); Est. of Hanges v. Metro. Prop. & Cas. Ins. Co., 202 N.J. 369 (2010); In re Est. of Jones, 259 N.J. 584 (2025)

The Court reaffirmed that where evidentiary admissibility is precedent to summary judgment, the evidentiary decision must be made first. Because expert admissibility was not properly resolved under Accutane, summary judgment could not properly be affirmed on the assumption of exclusion.

9) PLA substantive backdrop: Myrlak v. Port Auth. of N.Y. & N.J., 157 N.J. 84 (1999); Navarro v. George Koch & Sons, Inc., 211 N.J. Super. 558 (App. Div. 1986); In re Reglan Litig., 226 N.J. 315 (2016); Perez v. Wyeth Labys., Inc., 161 N.J. 1 (1999)

While the decision is primarily evidentiary, the Court located the dispute in the causation element of manufacturing defect and failure-to-warn claims under the PLA. These cases were used to explain the defect frameworks (manufacturing deviation; warning adequacy) and the “learned intermediary rule” statutory scheme.

10) Appellate issue preservation: State v. Robinson, 200 N.J. 1 (2009) and Nieder v. Royal Indem. Ins. Co., 62 N.J. 229 (1973); plus footnote reference to Coffman v. Keene Corp., 133 N.J. 581 (1993)

In a footnote, the Court declined to reach a heeding-presumption/proximate-cause argument because it was not properly presented below. This reinforces that even in high-stakes products cases, appellate courts will generally enforce issue-preservation rules.

B. Legal Reasoning

1) The core holding: Who decides reliability first—and how

The Court’s central doctrinal move is institutional: when reliability is contested under N.J.R.E. 702 and 703, the trial court must do the first-instance Accutane gatekeeping analysis. Appellate courts do not substitute themselves for that process on a thin record; their task is to review the trial court’s gatekeeping ruling (often for abuse of discretion) once a proper record exists.

2) Differential diagnosis is not exempt from Accutane

The Court rejected the “Creanga-only” approach. It clarified that differential diagnosis is not a shortcut around reliability scrutiny. Instead:

  • “Rule in” step: The expert’s inclusion of each “plausible cause” must itself be methodologically supported; ruling in a cause that is not “generally capable” of producing the injury is unreliable.
  • “Rule out” step: The expert must explain why alternative causes are eliminated using scientific methods and procedures, not mere temporal association or ipse dixit.

In practical terms, the Court required trial courts to analyze the differential diagnosis as a chain of methodological decisions, rather than accepting the label “differential diagnosis” as self-validating.

3) Reliability vs. net opinion: doctrinal separation matters

The Appellate Division treated shortcomings in proof and methodology as net opinion defects and independently concluded the opinions would not satisfy Daubert. The Supreme Court separated those doctrines:

  • Net opinion asks whether the expert supplied the “why and wherefore” and identified factual bases and methodology.
  • Accutane reliability asks whether the methodology and underlying data are sufficiently reliable to be presented to a jury.

Thus, Dr. Lalezary’s opinion could avoid being a net opinion (because it was explained) while still being subject to exclusion if, after a rigorous Accutane review, the methodology is found unreliable.

4) Treating physicians and written reports: compliance is not optional

A notable procedural clarification is that a treating physician who will give expert causation testimony must furnish a report when requested by interrogatory under Rule 4:10-2(d) and Rule 4:17-4(a), (e). The Court cautioned that admissibility determinations should be made with the expert report in hand and remanded for the trial court to decide whether a late report should be allowed and, if so, how defendant may respond.

5) Summary judgment depends on the evidentiary ruling

Because the admissibility of expert causation proof is often dispositive in pharmaceutical/product cases, the Court reaffirmed the sequencing principle: courts must decide admissibility first, then assess whether the remaining record supports summary judgment.

C. Impact

1) Trial-court practice: more robust gatekeeping (and more Rule 104 hearings)

By insisting that Accutane governs “any dispute” about reliability in civil cases and “strongly encourag[ing]” a N.J.R.E. 104 hearing on remand, the Court signals that litigants should expect a structured, record-based pretrial reliability process—especially in complex causation disputes involving medicine, pharmaceuticals, or epidemiologic/scientific inference.

2) Differential diagnosis testimony: heightened scrutiny of each inferential step

The decision will influence how parties prepare medical causation cases: experts must be prepared to defend (a) why each candidate cause is scientifically plausible (“rule in”) and (b) why alternatives are excluded (“rule out”) using transparent, scientifically grounded reasoning. Mere temporal correlation (“it happened after”) or reliance on recall status alone may be insufficient once subjected to the required Accutane scrutiny.

3) Appellate discipline: reliability disputes belong in the trial court first

The Court’s reversal of the Appellate Division underscores a process rule with substantive consequences: appellate courts should not resolve reliability in the first instance without a developed Accutane record and findings. This will likely produce more remands when trial courts skip gatekeeping steps.

4) Treating physician disclosures: fewer “deposition-only” expert designations

The Court’s reminder about treating-physician report obligations may curtail a common tactic of relying on depositions in lieu of reports, and it provides defendants stronger grounds to compel compliance and to challenge admissibility where disclosures are inadequate.

4. Complex Concepts Simplified

  • N.J.R.E. 702 / 703: The rules governing whether expert opinions can be presented (702) and what facts/data an expert may rely on (703).
  • Gatekeeping (Accutane): The trial judge must screen expert opinions for methodological reliability before the jury hears them.
  • Daubert factors (as used in New Jersey): Non-exclusive considerations such as testability, peer review/publication, error rate/standards, and general acceptance.
  • Differential diagnosis (legal sense): A two-stage reasoning process: (1) identify all plausible causes (“rule in”); (2) eliminate causes to reach the most likely cause (“rule out”).
  • General vs. specific causation: General causation asks whether an agent can cause an injury at all; specific causation asks whether it did so in the plaintiff’s case. The opinion emphasizes reliability scrutiny at both stages embedded in “rule in/rule out.”
  • Net opinion: An expert conclusion with no explained basis—no “why and wherefore.” It is excluded because it invites speculation.
  • N.J.R.E. 104 hearing: A pretrial evidentiary hearing where the judge decides admissibility questions, often used for expert reliability disputes.

5. Conclusion

Beavan v. Allergan establishes a clear, broadly applicable procedural and methodological directive: when expert reliability is challenged in a civil case, the trial court must conduct the Accutane “rigorous” gatekeeping inquiry, and that obligation squarely includes experts who rely on differential diagnosis—requiring scrutiny of both the “rule in” and “rule out” steps. The decision also reinforces disclosure discipline for treating physicians expected to provide expert causation testimony and reaffirms that summary judgment should be revisited only after admissibility is properly decided on a developed record.