The Brandywine Review

Commentary on American law

Health and End of Life Law

Health and End of Life Law

A federal judge in California spent ten days in 2017 hearing testimony about how an insurer wrote its own medical necessity guidelines, and concluded afterward that United Behavioral Health had built rules stingier than what psychiatry itself treated as generally accepted care. That finding, from the bench trial underlying Wit v. United Behavioral Health, ordered UBH to reprocess thousands of denied claims and was treated for several years as the clearest statement yet that federal mental health parity law had teeth. The Ninth Circuit disagreed. Reviewing the case for abuse of discretion under ERISA rather than deciding the clinical question itself, the panel held that UBH's own reading of its plans, that they did not require consistency with generally accepted standards of care, was not unreasonable, and it then revised its own reasoning twice more over the following year and a half. Wit v. United Behavioral Health, 79 F.4th 1068 (9th Cir. 2023), vacated the January opinion in full and sent the remaining fiduciary duty and exhaustion questions back to the district court.

The Mental Health Parity and Addiction Equity Act, 29 U.S.C. ยง 1185a, requires that treatment limitations on mental health and substance use benefits be no more restrictive than those applied to medical and surgical benefits. What the Wit litigation shows is that the statute's force depends almost entirely on which standard of review a court applies to an insurer's own interpretation of its own plan, and ERISA's deferential standard was built for pension disputes, not for clinical judgment calls about what generally accepted psychiatric practice requires. A parity statute enforced through deferential review of the defendant's own guidelines is a considerably weaker promise than the same statute enforced by a court asking directly whether those guidelines match accepted practice, and after three rounds of appellate opinions in one case, which standard actually governs remains genuinely open.

Civil commitment law has close to the opposite problem: the constitutional floor is fixed, but the words each state uses to describe who falls below it are not. Addington v. Texas, 441 U.S. 418 (1979), set that floor, holding that due process requires clear and convincing evidence, a standard well above the ordinary civil preponderance, before a state may commit someone involuntarily. Kansas v. Hendricks, 521 U.S. 346 (1997), confirmed that a state may go further still and commit a person indefinitely after a criminal sentence ends, so long as the statute ties commitment to a mental abnormality that makes future dangerousness likely rather than to the past offense alone. Neither case tells a state how to define dangerousness, and many commitment statutes still use language drafted decades before the diagnostic categories clinicians rely on today existed.

Someone who would be committed under one state's statute walks free under a neighboring state's, not because the clinical picture differs but because the words on the two states' books do.

The end of life side of the docket runs through a different due process case entirely. Cruzan v. Director, Missouri Department of Health, 497 U.S. 261 (1990), assumed that a competent person holds a constitutional liberty interest in refusing treatment, but held that a state may require clear and convincing evidence of an incompetent patient's own wishes before her family may withdraw life support on her behalf. That evidentiary bar is exactly what most families cannot meet without paperwork most people never complete. In re Quinlan, 70 N.J. 10 (1976), the case that opened this area of law fourteen years before Cruzan, grounded the same right to refuse treatment in the federal constitutional right to privacy running through Griswold and Roe, adding only that "[n]or is such right of privacy forgotten in the New Jersey Constitution." Nearly every state has since built some form of advance directive statute or surrogate decision hierarchy in Quinlan's wake.

Every state now lets a competent adult name a healthcare agent in advance, and every state also keeps a default hierarchy of who decides when nobody did. Where the statutes are weaker is the case in between: the directive that exists but was drafted for a different kind of decline than the one the patient actually faces, or the surrogate with clear legal authority and no real idea what the patient would have wanted about a treatment nobody ever discussed with her. Cruzan protects the right to have a patient's wishes honored once they are known. It does not reach the far more common situation of wishes that were never specific enough to be honored in the first place, and no later case has closed that gap.