Hook
A Notebook Kept for One Purpose
A therapist keeps a spiral notebook for exactly one purpose: proving, on command, that a person is still sick enough to deserve help.
On January 30, 2024, a Texas therapist opened that notebook and wrote three things at the top of a blank page — a date, a patient’s initials, and the name of a stranger she was about to call.1 The stranger was a psychiatrist employed by Blue Cross and Blue Shield of Texas. He would decide, in the space of one phone call, whether her patient — a woman named Geneva Moore, working through suicidal ideation, self-harm urges, and a long trauma history in an intensive outpatient program — got to keep her treatment.
Two weeks earlier, a different BCBS doctor had already said no to a colleague making the same case. The therapist had one shot left. She spent three hours before the call combing through Moore’s weekly assessments, her group session notes, everything that could show the treatment was working. Then she got on the phone and spoke as fast as she could.
It didn’t matter. The insurance company’s letter, when it came, gave its reason in four words: “You have made progress.”2
That phrase is the whole story. Moore wasn’t denied care because she was fine. She was denied care because she was getting better — carefully, incompletely, dangerously reversibly better — and someone with a job title and a phone line decided that improvement was itself the evidence that she no longer needed the thing that was producing it.
It’s the same shape of confidence this series traced not long ago — a wrongful conviction undone only after twenty-seven years on death row, once evidence nobody had been required to check was finally reviewed. One credentialed voice, speaking with total certainty, never made to sit in the room and watch what actually happens next. That story took decades to unravel in a courtroom. This one plays out in a few minutes on hold, in inboxes, in therapists’ notebooks, thousands of times a week, and almost nobody ever hears about it at all.
Historical Case
The Ward That Had to Look Curable
Long before insurance companies existed in their modern form, American institutions had already discovered that a patient’s appearance of improvement could be more useful to the organization than the patient’s actual recovery.
In the mid-1800s, the American asylum movement was built on a promise: with proper “moral treatment,” a fresh building, orderly grounds, and a firm daily schedule, most forms of insanity could be cured. Superintendents published annual reports boasting cure rates of seventy, eighty, even ninety percent — numbers that funded new construction and justified state appropriations.3 In 1877, the psychiatrist Pliny Earle published a landmark study showing that many of those figures had been produced by counting the same patients as “cured” more than once across repeated admissions, and that true recovery rates were far lower than institutions had spent decades claiming.4
Asylums that depended on demonstrating high cure rates had every incentive to admit patients who looked likely to recover quickly, discharge them the moment they showed enough improvement to be counted as “cured,” and quietly decline the harder, more chronic cases that would drag the statistics down. A woman who was improving was, paradoxically, at greater risk of being sent home too soon than a woman who showed no improvement at all — because her progress had become useful to someone else’s bookkeeping.
Why would an organization built to heal people ever have a reason to rush that healing?
It’s worth being precise about what the two systems actually share, because it isn’t their motive. The nineteenth-century asylum needed public trust and state appropriations to keep its doors open; it was, at least in principle, answerable to a reputation. The modern insurer needs a favorable loss ratio to satisfy shareholders and regulators; it is answerable to a margin. One was a public institution protecting its standing. The other is a private enterprise protecting its bottom line. What they share isn’t the incentive — it’s the mechanism that incentive produces: an organization that measures its own success by the same numbers it uses to justify letting a person go.
Nothing about Geneva Moore’s case was new. It was the oldest trick a caregiving system can play on a person healing inside it: measuring their progress not by what they still need, but by what the measurer would prefer to stop paying for.
Biblical Lens
The Bible is unusually attentive to the difference between healing that has started and healing that is finished — and to how dangerous it is to confuse the two. Mark’s Gospel records a moment of exactly this ambiguity, when Jesus heals a blind man in two stages rather than one:
“He looked up and said, ‘I see people; they look like trees walking around.’ Once more Jesus put his hands on the man’s eyes. Then his eyes were opened, his sight was restored, and he saw everything clearly.” (Mark 8:24–25)
Readers familiar with Mark’s structure sometimes point out that this account sits directly before Peter’s confession at Caesarea Philippi, and read the two-stage healing primarily as a literary hinge — a picture of the disciples’ own gradual spiritual sight rather than a statement about medical process as such. That reading has real merit, and this essay doesn’t need to dispute it to make its point. Even taken purely as literary structure, the text still declines to let the in-between state stand in for the finished one. Whatever else the two stages are doing, Jesus does not stop at “trees walking around.” He returns, a second time, and continues until the man sees “everything clearly” — not until the man reports enough improvement to justify moving on. The healer’s hands stay on the patient until the healing is actually finished, not until it is merely underway.
Paul’s language to the believers in Philippi carries the same logic forward, applied to the whole shape of a life rather than a single healing: “he who began a good work in you will carry it on to completion until the day of Christ Jesus” (Philippians 1:6). The responsibility for finishing belongs to the one who started the work — not to whoever happens to be watching from outside and deciding when it’s convenient to call the project done.
And Proverbs adds a warning aimed specifically at the party doing the withholding, not the party left wanting: “One person gives freely, yet gains even more; another withholds unduly, but comes to poverty” (Proverbs 11:24). Israel’s wisdom tradition treated premature withholding as a form of poverty-in-the-making — a warning better aimed at institutions than at patients.
None of this is a claim that every insurance denial is a spiritual failing, or that every insurer acts in bad faith. The improvement was proof the process was working, not proof it was over — and Scripture’s oldest healing accounts already knew the difference, even when the healer himself chose not to stop at the first sign of it.
Pattern Insight
What the Numbers Confirm
Geneva Moore’s case was never an outlier. It was one well-documented instance of a pattern that shows up in the data at a scale that’s difficult to absorb in the abstract.
According to the American Psychiatric Association’s 2024 parity analysis, behavioral health claims are denied at a rate roughly 85 percent higher than comparable medical claims — this despite a federal law, the Mental Health Parity and Addiction Equity Act, that has explicitly required equal treatment of the two since 2008.5 In Georgia alone, state regulators logged more than six thousand parity violations across twenty-two insurers in a recent enforcement sweep spanning 2025 and 2026 — the largest state action of its kind since the federal law was passed.6
It’s worth pausing to say plainly what insurers themselves argue, because the argument isn’t absurd on its face: utilization review exists to prevent unnecessary treatment and to preserve limited clinical resources for the patients whose need is most acute. That purpose is legitimate, and no honest account of this system should pretend otherwise.
But the same insurance industry that denies Geneva Moore’s care on the basis of improvement does not apply that logic elsewhere in medicine. No physician manages diabetes this way. When a patient’s blood sugar responds to insulin and begins to normalize, no doctor — and no insurer — treats that improvement as grounds to withdraw the insulin. Improvement is read, correctly, as evidence the treatment is working and should continue until the underlying condition is stable, not merely improving. Behavioral health is governed by a different, unwritten rule, in which improvement itself becomes the reason care stops. The clinical logic is identical in both cases. Only the diagnosis category changes the response — and that asymmetry is precisely what parity law was written to eliminate, and precisely what the denial-rate data suggests it has not.
What happens to the people behind the six thousand violations once the appeal is denied?
Some get better anyway, on their own, at greater cost and greater risk. Some don’t. Researchers who track denial data have found that when mental health claims are formally appealed, the original decision is overturned more often than not — in California, in more than seven cases out of ten.7 That single statistic does something the raw denial numbers can’t: it shows the first call was frequently, provably wrong — discovered only after a person like Moore had already lost the momentum, the continuity of care, or in some documented cases, the window in which recovery could continue uninterrupted.
The deeper pattern here isn’t unique to insurance. Any organization that measures its own success by visible endpoints will eventually reward the endpoint over the continuation — because endpoints are what get counted, and continuation is what gets defunded. Schools graduate students whether or not the learning is finished. Courts close cases whether or not justice fully lands. Hospitals discharge patients whether or not the healing outlasts the bed they were occupying. Bureaucracies of every kind are built to recognize completion far more easily than they are built to recognize a process still quietly underway — and a person’s own improvement, ironically, is the clearest completion signal available to an organization looking for a defensible place to stop.
The asylum superintendents of the 1800s needed good numbers to keep their doors open. Today’s insurers need favorable ratios to satisfy the same kind of structural pressure, dressed in modern vocabulary — utilization review, medical necessity, evidence-based discharge criteria. The words changed. What a system does with a patient’s own progress did not.
Closing
What the Notebook Couldn’t Change
Geneva Moore’s therapist lost that call. The denial stood. What happened to Moore in the weeks after is not part of the record most people will ever read — most stories like hers end exactly there, in a letter, in a closed file, in a number that helped somebody’s quarterly report look a little better than it otherwise would have.
That’s the room I keep returning to — not the hospital room, not even the phone call, but the one just before it: a therapist alone with a notebook, gathering every piece of evidence a human being’s healing had ever produced, preparing to lay it in front of someone who had already decided what counts as enough.
Institutions rarely collapse because they become incapable of measuring outcomes. They fail when they begin mistaking the measurement for the person the measurement was invented to serve.
The asylum ledgers of the 1800s and the denial letter Geneva Moore received a century and a half later were never really about medicine, or money. They were about which came first — the person, or the proof.
1. ProPublica, “Mental Health Coverage Denied? It May Be Because Your Insurer Said You Made Too Much or Too Little ‘Progress,'” Jan. 3, 2025.
2. Ibid.
3. Gerald N. Grob, Mental Illness and American Society, 1875–1940 (Princeton University Press, 1983).
4. Pliny Earle, The Curability of Insanity: A Series of Studies (J.B. Lippincott, 1887); see also Grob (1983) on asylum statistical practices.
5. American Psychiatric Association, 2024 Mental Health Parity Report.
6. Georgia Office of Insurance and Safety Fire Commissioner, enforcement data spanning 2025–2026, as reported by The Kennedy Forum and MoneyGeek, 2026.
7. California Department of Managed Health Care appeals data, as reported 2023.
New essays every week. Delivered quietly.