Abuse-of-Discretion, Not “Arbitrary and Capricious”: Sixth Circuit Clarifies ERISA Review and Confirms New Physician File Reviews Can Justify LTD Denials After STD Approvals

Abuse-of-Discretion, Not “Arbitrary and Capricious”: Sixth Circuit Clarifies ERISA Review and Confirms New Physician File Reviews Can Justify LTD Denials After STD Approvals

Introduction

In Brandi Goodwin v. Unum Life Insurance Company of America, No. 24-3321 (6th Cir. May 15, 2025), the Sixth Circuit affirmed the denial of long-term disability (LTD) benefits under an ERISA-governed plan. Writing for the court, Judge Thapar clarified the standard of review the circuit applies when a plan grants discretionary authority to an administrator: although the court’s cases have often referred to “arbitrary and capricious” review, the proper label—consistent with Supreme Court trust-law guidance—is “abuse of discretion.” The court also set out a structured framework for evaluating an administrator’s decision on both procedural and substantive reasonableness and held that “new evidence” sufficient to justify a change from approving short-term disability (STD) to denying LTD can include new professional medical file reviews, even where no “new raw medical data” has been generated for the same time period.

The case arises from Goodwin’s COVID-19-related symptoms and subsequent diagnoses (including a later POTS diagnosis). Unum approved her STD claim through the 24-week maximum but denied LTD. The district court upheld the denial, and the Sixth Circuit affirmed, concluding that Unum’s process was procedurally sound and its decision substantively supported by substantial evidence within the administrative record.

Summary of the Opinion

The Sixth Circuit affirmed the district court’s judgment upholding Unum’s denial of LTD benefits. The court’s key holdings are:

  • Standard of Review: Where the ERISA plan delegates discretionary authority, courts apply abuse-of-discretion review (not “arbitrary and capricious,” though the Sixth Circuit’s past usage of that term is functionally equivalent). The court expressly favored using the Supreme Court’s trust-law label of “abuse of discretion.”
  • Procedural Reasonableness: Unum’s process was reasonable. It considered all relevant evidence, adequately justified the change from STD approval to LTD denial by pointing to new evidence—including new physician file reviews—did not improperly elevate file reviewers over treating providers, and showed no conflict-of-interest taint supported by concrete record evidence.
  • Substantive Reasonableness (Substantial Evidence): Unum’s decision was supported by substantial evidence: normal tests and exams, documented activity and exercise tolerance, and multiple medical reviewers’ opinions concluding that disability was not supported after February 23, 2021.

The court declined to reach Unum’s initial “pre-existing condition” rationale because Unum did not rely on it on appeal. The court also noted, in dicta, concerns about the atextual use of “judgment on the administrative record” in ERISA litigation, while acknowledging that no challenge to that procedural practice was before the court.

Detailed Analysis

1) Precedents Cited and Their Influence

  • Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101 (1989) and Metropolitan Life Insurance Co. v. Glenn, 554 U.S. 105 (2008): The Supreme Court grounds ERISA review in trust law. Where a plan grants discretion to a fiduciary, courts review for abuse of discretion. Glenn further instructs courts to consider structural conflicts of interest as a factor. The Sixth Circuit here doubles down on Firestone/Glenn’s trust-law framework and consciously adopts the “abuse of discretion” label as the more accurate descriptor for review of private ERISA fiduciaries.
  • Autran v. Procter & Gamble Health & Long-Term Disability Benefit Plan, 27 F.4th 405 (6th Cir. 2022): Provides the procedural/substantive framework for evaluating reasonableness: courts ask whether the administrator engaged in reasoned decisionmaking (procedural) and whether substantial evidence supports the outcome (substantive). Autran also emphasizes that new evidence can justify a change in benefits determinations and requires “concrete evidence” to give significant weight to structural conflicts.
  • Wilkins v. Baptist Healthcare Sys., Inc., 150 F.3d 609 (6th Cir. 1998): Confirms de novo review of the district court’s judgment on the administrative record, with deference determined by plan terms.
  • Price v. Board of Trustees of Indiana Laborer’s Pension Fund, 632 F.3d 288 (6th Cir. 2011) and Cox v. Standard Insurance Co., 585 F.3d 295 (6th Cir. 2009): Illustrate the circuit’s historic toggling between “arbitrary and capricious” and “abuse of discretion” language; the court explains that, functionally, its approach has long centered on reasonableness consistent with trust law.
  • Black & Decker Disability Plan v. Nord, 538 U.S. 822 (2003): Rejects a special “treating physician rule” in ERISA; administrators need not give treating physicians’ opinions controlling weight or provide heightened explanations for preferring non-treating experts. The court relies on Nord to reject the argument that Unum erred in crediting file reviewers over treating clinicians.
  • McDonald v. Western–Southern Life Insurance Co., 347 F.3d 161 (6th Cir. 2003) and Bruton v. American United Life Insurance Corp., 798 F. App’x 894 (6th Cir. 2020): Reinforce that administrators do not act unreasonably merely by preferring one set of medical opinions over another, and courts do not accord special weight to treaters.
  • Morris v. American Electric Power Long-Term Disability Plan, 399 F. App’x 978 (6th Cir. 2010) and Holden v. Unum Life Insurance Co. of America, No. 20-6318, 2021 WL 2836624 (6th Cir. July 8, 2021): Establish that administrators may change their determinations between STD and LTD stages if they have a reason, and that it would be illogical to bind LTD decisions to prior STD approvals. Holden is quoted to underscore that requiring lockstep outcomes would collapse the distinction between STD and LTD and could deter prompt STD payments.
  • Eberle v. American Electric Power System Long-Term Disability Plan, No. 21-4224, 2022 WL 5434559 (6th Cir. Oct. 7, 2022): Demonstrates that file reviewers’ opinions can constitute substantial evidence supporting an administrator’s decision.
  • Evans v. Eaton Corp. Long Term Disability Plan, 514 F.3d 315 (4th Cir. 2008): Cited for the observation that “abuse of discretion” is arguably less deferential than “arbitrary and capricious”; used to frame why labels can matter, even if the Sixth Circuit’s functional analysis remains grounded in reasonableness.
  • Tranbarger v. Lincoln Life & Annuity Co. of N.Y., 68 F.4th 311 (6th Cir. 2023) (Nalbandian, J., concurring): Quoted in dicta questioning the atextual practice of ERISA “motions for judgment on the administrative record.” The court flags the issue but leaves it unresolved here.

2) The Court’s Legal Reasoning

a) Standard of Review: Embracing “Abuse of Discretion”

The plan undisputedly delegated discretionary authority to Unum. Under Firestone and Glenn, that triggers abuse-of-discretion review grounded in trust law. The Sixth Circuit acknowledges its historic habit of calling this “arbitrary and capricious” review (a term more commonly associated with administrative law and federal agency action) and clarifies that the correct label is “abuse of discretion.” The court emphasizes substance over form—its past practice has focused on reasonableness consistent with trust principles—but states that “labels affect thought,” and moving forward, it is best to use the Supreme Court’s trust-law nomenclature.

b) Procedural Reasonableness

Applying Autran, the court evaluates four non-dispositive factors:

  • Consideration of All Relevant Evidence. Unum and its reviewers addressed Goodwin’s POTS diagnosis, stress test data, reports of tachycardia, chest pain, and shortness of breath. Far from cherry-picking, Unum acknowledged those reports but weighed them against normal imaging, pulmonary function tests, consistently normal vitals and exams, clinical encouragement to exercise and return to work, and physical therapy notes documenting significant functional activity. The court rejected the argument that Unum ignored vertigo; Dr. Winkel accounted for it and noted absence of balance issues and dizziness during activities.
  • Explanation for Changing from STD Approval to LTD Denial. Under Autran’s “low bar,” administrators must offer a rational reason for a changed outcome; “new evidence” suffices. Here, the “new evidence” was not merely new raw medical data, but also new expert interpretive assessments: Unum’s LTD process included physician file reviews (Bress, Palermo, Winkel) that were not part of the STD approvals. Those new professional opinions—combined with later-year clinical records showing improvement and normal exams—amply justified the shift. The court also cites Holden to reject any rule that LTD must follow STD approvals.
  • Treating Providers vs. File Reviewers. Although some treaters (Drs. Ruggles and Cherian) supported disability, others (NP Gast) encouraged a return to work. In any event, Nord forecloses any presumption in favor of treaters. The administrator may reasonably credit non-examining specialists where their conclusions are grounded in the record.
  • Conflict of Interest. Unum’s dual role as adjudicator and payor creates a structural conflict, but Autran requires “concrete evidence” that the conflict tainted the decision. Goodwin identified no such record evidence. Thus, while the conflict is a factor, it carries little weight here and cannot overcome the other indicia of procedural reasonableness.

c) Substantive Reasonableness (Substantial Evidence)

The court asks whether “a rational person could conclude that the evidence was adequate to justify the decision.” On this record, multiple strands of evidence support Unum’s denial:

  • Normal pulmonary function testing and chest imaging,
  • Repeated normal vitals and exams, including normal oxygen saturation and absence of distress,
  • Documented exercise capacity (walking daily, jogging every other day, 30-minute weight training),
  • Clinical recommendations to return to work and engage in exercise, and
  • Convergent medical opinions from multiple file reviewers, including an outside consultant (Dr. Winkel), concluding disability was not supported after February 23, 2021.

Because the record could support a rational decision either way, Unum’s choice among competing medical views is substantively reasonable under the substantial-evidence standard.

3) Impact and Practical Implications

a) Terminology and Framing: Abuse-of-Discretion Going Forward

  • The court’s express preference for “abuse of discretion” aligns circuit practice with Supreme Court trust-law doctrine and may influence how district courts and litigants frame arguments and orders.
  • Expect increased emphasis on reasonableness and trust-law principles rather than administrative-law analogies when evaluating ERISA fiduciary determinations.

b) STD vs. LTD: “New Evidence” Includes New Expert File Reviews

  • The opinion confirms that administrators may reach different outcomes at different benefit stages. A key clarification: “new evidence” justifying a change can be new professional interpretations or file reviews added during the LTD stage—even if no new raw clinical tests exist for the overlapping time period.
  • Claimants cannot leverage STD approvals as presumptive entitlement to LTD. Administrators need not be “locked into” earlier conclusions if later reviews introduce new expert analysis.

c) Treating Physician Opinions

  • Nord remains dispositive: no special deference to treating opinions. Administrators may reasonably prefer well-explained file-review opinions, especially where clinical exams and objective testing are normal or show improvement.

d) Conflicts of Interest: Evidence Matters

  • Structural conflicts are ubiquitous and, without concrete record evidence of actual bias or process irregularities, will not tip the scale. Claimants should develop specific proof tying conflicts to decisionmaking (e.g., inconsistent application of standards, selective record development, or pressure on reviewers).

e) Procedural Posture: The Court’s Dicta on ERISA Procedure

  • The court signals skepticism about “motions for judgment on the administrative record” as atextual under the Federal Rules of Civil Procedure, but leaves the issue for another day. Litigants should watch for potential en banc or Supreme Court guidance, and consider preservation of objections to non-Rule-based procedures in ERISA litigation.

Complex Concepts Simplified

  • ERISA and Fiduciaries: ERISA regulates employer-sponsored benefit plans. Plan administrators are “fiduciaries,” meaning they must act in participants’ interests and in accordance with plan terms.
  • Delegation of Discretion: Many plans grant administrators discretion to interpret plan terms and make benefit determinations. When they do, courts review decisions for “abuse of discretion,” asking whether the decision is reasonable, not whether the court would have reached the same conclusion.
  • Procedural vs. Substantive Reasonableness: - Procedural reasonableness looks at how the decision was made (Was all relevant evidence considered? Was a change explained? Was there bias?). - Substantive reasonableness looks at whether enough evidence supports the outcome (Would a rational person find the evidence adequate?).
  • Substantial Evidence: Not the most evidence, but enough that a reasonable person could agree with the decision. If the record supports either of two outcomes, courts defer to the administrator’s choice under abuse-of-discretion review.
  • Structural Conflict of Interest: When the decisionmaker also pays benefits, there’s a potential conflict. It’s a factor in review, but it only carries significant weight when there’s concrete proof that the conflict influenced the decision.
  • STD vs. LTD: Short-term and long-term disability benefits are distinct benefits with distinct evaluations. An STD approval doesn’t guarantee LTD. Later-stage reviews may incorporate new expert opinions that justify a different outcome.
  • “Administrative Record” in ERISA: Courts typically confine review to the record compiled during the plan’s administrative process. Although the Sixth Circuit notes procedural concerns about the term “judgment on the administrative record,” that practice remains in place absent further guidance.

Conclusion

Goodwin v. Unum does not change the outcome-determinative principles of ERISA review so much as it clarifies and sharpens them. The Sixth Circuit endorses the abuse-of-discretion label for cases involving discretionary plan authority, reaffirming that reasonableness—grounded in trust law—governs both procedure and substance. Administrators may change course from STD to LTD where “new evidence” exists, and crucially, the court recognizes that new, professional physician file reviews can constitute that new evidence even in the absence of new raw clinical data for the same time frame.

The opinion also reiterates the rule that treating physicians receive no special deference under ERISA and that structural conflicts will carry little weight absent concrete proof of taint. Finally, the court’s dicta questioning the atextual “judgment on the administrative record” signals possible future refinement of ERISA procedure in the circuit.

Key takeaways:

  • Expect courts and litigants in the Sixth Circuit to use “abuse of discretion” to describe the standard where plans confer discretion.
  • STD approvals do not predetermine LTD outcomes; new expert reviews can justify a different result.
  • Substantial evidence can include normal testing, documented functional activity, and convergent file-review opinions.
  • Conflict-of-interest arguments must be supported by specific record evidence to gain traction.
  • Watch for potential procedural developments regarding ERISA’s use of “judgment on the administrative record.”

On the facts presented, Unum’s denial survived both procedural and substantive scrutiny. For future ERISA disputes, the opinion provides a clear template: build the record, engage specialists where appropriate, explain changes with reference to new evidence (including new expert analyses), and ensure the file reflects reasoned decisionmaking throughout.

Case Details

Year: 2025
Court: Court of Appeals for the Sixth Circuit

Comments