Poor insight.
Two words that fit easily into a psychiatric chart. Sometimes they aren’t even a sentence you write. They’re a choice in the mental status exam: good, fair, poor. You pick one and keep going, because there are eleven more fields and the shift is what it is.
I’ve picked poor. I picked it the same way I’ve written “cooperative with care,” because it was the word available for something I was seeing and the form had a slot for it.
I can’t trace the language in A.B.’s case back to a nurse’s chart. But I can follow it from the moment a psychiatrist said it on the stand.
The case
In November 2001, a man the Indiana courts call A.B. drove to a mall in Clarksville. He’d been hearing voices for about a year. They warned that people were entering his house, that he was being set up for a crime that would let police kill him. He boarded up his windows. He carried a gun and slept with it.
In the parking lot he saw a man he’d met on vacation in Jamaica. The voices told him this man was setting him up, that the man’s brother was a police officer who’d plant drugs in his house, that he could be killed and his death staged as a suicide. A.B. pulled the man from his car and shot him in front of his wife.
A jury found him not guilty by reason of insanity. In February 2002 the trial court committed him to Madison State Hospital, and the commitment kept getting renewed. In late 2009 he requested review. After four hearing dates between June 2010 and February 2011, the court continued the commitment. The Court of Appeals affirmed in July 2012.
I’m telling you the killing first because you’d find it in thirty seconds and then wonder what else got left out.
The findings
Indiana required clear and convincing evidence of mental illness and either dangerousness or grave disability. Grave disability required that the person be in danger of coming to harm because of the illness, and then supplied two branches: inability to provide for essential needs, or substantial impairment or obvious deterioration of judgment, reasoning, or behavior resulting in inability to function independently.
Both parts of that second branch have to be there. Impaired judgment, and an inability to function independently that results from it.
The trial court found both dangerousness and grave disability, using the independent-functioning branch. Fifty-three findings of fact, thirty-four conclusions of law, all reproduced in the appellate opinion.
Finding 38 says A.B. is gravely disabled because he lacks insight into his illness and exercises poor judgment about community stressors and triggers. Finding 39 treats his wish to return to the home and community where the killing happened as evidence of that.
Then Finding 40, substantive text, transcript citations omitted:
Although [A.B.] and his family acknowledge that he suffers from schizophrenia, that this is a lifelong chronic illness, and that he will require medication for the rest of his life, they do not appreciate the significance of his illness and the potential impact of stressors and triggers in the community. Thus, [A.B.], and his family are in denial about important aspects of [A.B.]’s mental illness.
He acknowledges the diagnosis. Acknowledges it’s lifelong. Acknowledges he’ll need medication the rest of his life. His family does too. And the finding is denial.
The finding distinguishes acknowledging an illness from appreciating what it might mean outside a hospital, and those really are different things. Finding 41 names an object: A.B. does not display acceptable insight into his mental illness and his risk of relapse. Dr. Porter, his treating psychiatrist, testified that A.B. had told him a few days earlier that he had no mental illness, that he’d said forcefully and in anger that he should never have been put in a psychiatric hospital, that he minimizes the illness when he isn’t denying it. Finding 42 says A.B. called the killing “a bad turn of events” and talked about coming out of it without medication or treatment.
Findings 30 and 33 add current concerns. Threatening outbursts. An encounter where Dr. Porter felt threatened. Verbal aggression toward a recreational therapist. An episode where A.B. said he began to feel as he had in 2001 after being denied access to a computer.
Finding 29 goes the other way. During his hospitalization, in a structured setting, on medication and in therapy, A.B. had not been violent. Finding 31 says that even so, no expert testified that he no longer posed a risk of danger to others.
All of that is in the opinion and all of it matters. Acknowledgment, reported denial, reported minimization, threatening incidents, and a hospital record without violence, and the opinion never resolves them into one account of what A.B. understood or what he could do.
The family finding is the weakest link and worth its own look. Finding 26 says A.B. had successfully masked his paranoia and psychosis before the killing and his family hadn’t noticed. That history could reasonably matter to a discharge plan built on family support. It doesn’t establish, on its own, what they understood eight years later. For that, the appellate court’s explanation rests on Dr. Porter relaying another physician’s concern that the parents were trying to stifle A.B.’s discussion of his psychosis or delusions or hallucinations, “something to that effect.” That’s the doctor’s wording as the opinion reproduces it. I haven’t read the full evidentiary record, so there may have been more. But a secondhand, approximately described concern is what the appellate court reached for when it explained why the finding about the family stands.
Insight into what
For some people with schizophrenia, impaired awareness of illness appears to be part of the illness rather than stubbornness or a bad attitude. Lehrer and Lorenz make that case for extending the anosognosia construct to schizophrenia, and it’s a real phenomenon that clinicians see. It’s also compatible with everything below, because a condition can be clinically real and still fail to prove what a statute requires.
The same literature shows that insight has parts. When Michel and colleagues validated an abbreviated version of the Scale to Assess Unawareness of Mental Disorder, one of the most widely used research instruments, in 531 patients, a three-factor structure held: awareness of the disease and its consequences and the need for treatment, awareness of positive symptoms, awareness of negative symptoms. Xavier Amador, who built the original scale, is a co-author. And a systematic review by Dumas and colleagues found that across a hundred studies over twenty years, researchers rarely used the scale in full and varied widely in how they scored and interpreted it, enough that the authors warn modified use can compromise its properties.
That says nothing about how Dr. Porter assessed A.B. What it supports is narrower. Insight is not one thing, and an assessment that uses the word should name which part it means and show what it rests on.
Refusal also can’t tell you its own cause. People stop medication because of side effects, because the pharmacy is forty minutes away, because coverage lapsed, because it made them feel dead. Insight does track with adherence and with a stronger working relationship with clinicians, at moderate strength, in a study of 903 European outpatients.
Finding 40 might be describing something real. It still has to be connected to whether A.B. could function independently.
The alternatives that didn’t exist
Indiana’s outpatient statute required four findings, and the fourth was a recommendation from the person’s examining physician. Dr. Porter testified that A.B. needed to be at Madison State and that outpatient wasn’t what he was recommending. The Court of Appeals held that defeated the fourth element and expressly declined to analyze the other three. On this record, one doctor’s non-recommendation ended the inquiry.
The practical picture is in the findings. LifeSpring, the community mental health center, had an Assertive Community Treatment team with no openings. No group home placement available. A.B. couldn’t enter its outpatient program immediately. And the ACT team didn’t serve people living in group homes, so those two unavailable things weren’t pieces of one plan anyway. Finding 51: Indiana had no conditional release program of the kind some states run. Finding 52: without a commitment order there was no statutory mechanism for court supervision or ensuring treatment.
The trial court put the resource problem in its conclusions of law. Conclusion 28 invokes separation of powers. Conclusion 29 says the court cannot order the state mental health agency to provide placements and services that do not exist or are not presently available. Conclusion 30 identifies Madison State Hospital as the least restrictive environment suitable for A.B.’s care.
Read those three in order. The court put its own constraint in writing, and then what Indiana had failed to build ended up as a sentence about A.B.
I don’t think money alone kept him hospitalized, and the opinion doesn’t establish that. There were clinical concerns, and we don’t know whether Dr. Porter would have recommended a well-supported community plan if one had existed. But what didn’t exist helped define the alternatives against which the hospital was judged.
I know this pressure from nursing. What you can do on a shift depends on how many people are working it. On a thin night the thing that might have helped isn’t available. The chart records the behavior. It has no field for the option that wasn’t there.
The footnote
A.B. argued the state owed him reasonable treatment to improve his condition, and that if lack of insight was really the barrier to discharge, Madison State had failed to recognize and treat it from 2003 to 2009, so continued confinement on that basis violated due process.
That’s an allegation and an alternative argument, not an established six-year treatment failure. He was also challenging the insight finding itself. We don’t have his treatment history or his full brief.
The court called the argument specious, said he hadn’t supported it with cogent reasoning, and declined to address it. That’s the entire discussion. It didn’t decide whether the hospital had failed or what would follow if it had.
A footnote won’t tell me whether the argument was developed well enough to deserve better. The question underneath it is still a fair one. If a particular impairment is the barrier to discharge, what treatment addresses it, what would improvement look like, and what happens if it persists despite good care? An impairment doesn’t stop being real because it’s hard to treat. But a finding that recurs year after year should come with an account of the care being offered and what would count as enough.
The standard
The Court of Appeals affirmed on grave disability. That was the ground it found sufficient, so it never decided whether A.B. was also dangerous.
When the court set out how it would review the evidence, it wrote that it looks only at the evidence most favorable to the judgment, citing J.S. v. Center for Behavioral Health, that it won’t reweigh evidence or judge credibility, and then this, from M.M. v. Clarian Health Partners: if the commitment order represents a conclusion a reasonable person could have drawn, the court will affirm even if other reasonable conclusions are possible.
Clear and convincing evidence appears nowhere in that paragraph.
In 2015 the Indiana Supreme Court decided T.K. v. Department of Veterans Affairs. Before it reached the merits it took up the standard of review, and said several recent Court of Appeals decisions had recited the clear and convincing language without applying it, affirming instead whenever the order represented a conclusion a reasonable person could have drawn. It named four of them. One is J.S. v. Center for Behavioral Health.
We disapprove of this line of cases, the court wrote. The question is whether the trial judge could have found by clear and convincing evidence that the person was dangerous or gravely disabled.
Then the merits. The psychiatrist had based his grave-disability opinion on T.K.’s refusal of treatment, denial of mental illness, and reports of aggression. The evidence showed loud and coarse behavior, screaming at clinic staff, and flyers left on windshields, which Dr. Bishara understood to have reproduced the public sex-offender record of the man who had molested T.K.’s daughter, the husband of T.K.’s ex-wife. It also showed that T.K. rented a home, lived alone, held a full-time job, and maintained his household and vehicles, and that he’d secured work and housing during six months off medication. The court held that denial of illness and refusal to medicate, standing alone, were insufficient to establish grave disability because they don’t establish an inability to function independently, that the offensive behavior didn’t close the gap, and that the evidence of dangerousness fell short too. Reversed.
A.B. was decided under the disapproved standard, and it cited one of the disapproved cases to get there.
T.K. doesn’t mention A.B., doesn’t decide it, and the two records aren’t comparable. It also kept the deference that comes with sufficiency review, still looking at the evidence and inferences favorable to the judgment, and it didn’t require any particular list of daily activities or rule out predictive clinical testimony. What it reaffirmed is that sufficiency review carries the heightened burden with it.
I’d be careful with T.K.’s facts too, because the house and the job and the vehicles can slide into a claim I don’t believe: that you earn your way out of a locked ward by holding a job and keeping your car running. Plenty of people can’t do those things and none of them belong in a hospital for it. What matters isn’t the contents of that list. It’s that somebody went and looked at what the man actually did with his days.
A.B.’s opinion has a different kind of evidence at its center. Alongside the testimony about denial and minimization, Dr. Porter told the court that if released, A.B. posed a risk of rapid decompensation and relapse in the community, and that he was still gravely disabled. The risk he described had two sources: leaving the structure of the hospital, and returning to a specific place, the town where the killing happened, the workplace he’d had before, the people still living there.
Predictions like that are legitimate clinical work. Anyone who has discharged a patient on a Friday has made one, and a well-supported forecast can be real evidence about whether someone can function outside a hospital. So the objection isn’t that Porter predicted. It’s what the prediction leaves open. Which capacities would fail. What would have to be in place to keep them from failing. Whether anything short of a locked ward would do, or whether the answer was simply that nothing else was available.
The opinion is not silent on the surrounding questions. It addresses whether hospitalization was necessary, whether outpatient care was recommended, what LifeSpring had open, and what Indiana had never built, and it addresses all of that at length. It also states the connection outright, holding that the evidence supports a determination that A.B. has a substantial impairment or obvious deterioration of judgment, reasoning, or behavior that results in his inability to function independently.
Stated, though, and not shown. The opinion does name an impairment: judgment about the stressors and triggers waiting in that community. What it doesn’t develop is the step from there to the statute. Denial, minimization, a predicted relapse, and then the conclusion that he cannot function independently, with the middle of that reasoning left out. A reader can examine the inference. They just aren’t given much to examine it with. That’s a complaint about the explanation, not a conclusion about what the evidence could have supported if someone had made the connection visible.
And what he could do would depend on what he was given. That’s where the ACT team with no openings comes back in. You can’t answer whether someone can function independently without saying what they’d be functioning with.
Charting
I don’t know what happened to A.B. after July 2012. The decision is a nonprecedential memorandum decision, and Indiana’s rule generally bars citing a decision of that vintage to a court except for specified uses. That restriction is about courtroom citation. It doesn’t stop anyone from reading it.
I still chart. I still have to write down what I see, and sometimes what I see is someone whose account of their own situation would get them hurt if we acted on it. That’s real. The assessment is not empty.
What changed is smaller and more irritating than a change of heart. When I get to that field I try to put the object back in, then the observation under it, then what the person had to work with, and I try to keep my theory separate from what the record shows.
An invented example, not a real patient: pharmacy reports delivery stopped in December when coverage lapsed; refill records show no fills in January; patient reports he stopped the medication because it made him feel flattened, and says the March admission was unrelated; nursing observed him filling his own weekly organizer in November.
That takes longer and it doesn’t fit the box. He may be right that the two things are unrelated. But the note puts the December lapse in the record beside the January gap, and it shows where each piece came from, so that anyone deciding what this man can do has to look at what he had to do it with.
Source note: This essay works from the appellate opinions and the trial-court findings and conclusions reproduced in them, not the complete trial record. Numbered findings are the trial court’s. The charting example is invented.
Cases: In the Matter of the Involuntary Commitment of A.B., No. 10A01-1109-MH-483 (Ind. Ct. App. July 3, 2012) (mem. dec.). In the Matter of the Civil Commitment of T.K. v. Department of Veterans Affairs, 27 N.E.3d 271 (Ind. 2015). The disapproved line in T.K. comprises M.L. v. Meridian Services, Inc., 956 N.E.2d 752, 755 (Ind. Ct. App. 2011); S.T. v. Community Hospital North, 930 N.E.2d 684, 688 (Ind. Ct. App. 2010); K.F. v. St. Vincent Hospital & Health Care Center, 909 N.E.2d 1063, 1066 (Ind. Ct. App. 2009); and J.S. v. Center for Behavioral Health, 846 N.E.2d 1106, 1111 (Ind. Ct. App. 2006).
Clinical sources, retrieved via PubMed: Douglas S. Lehrer & Jennifer Lorenz, Anosognosia in Schizophrenia: Hidden in Plain Sight, 11 Innovations in Clinical Neuroscience 10 (2014), PMC4140620 (review and conceptual argument, not new data). Pierre Michel et al., Psychometric Properties of the Abbreviated Version of the Scale to Assess Unawareness in Mental Disorder in Schizophrenia, 13 BMC Psychiatry 229 (2013), doi:10.1186/1471-244X-13-229. Rémy Dumas et al., Systematic Review Reveals Heterogeneity in the Use of the Scale to Assess Unawareness of Mental Disorder (SUMD), 15 Current Psychiatry Reports 361 (2013), doi:10.1007/s11920-013-0361-8. Diego Novick et al., Relationship of Insight with Medication Adherence and the Impact on Outcomes in Patients with Schizophrenia and Bipolar Disorder, 15 BMC Psychiatry 189 (2015), doi:10.1186/s12888-015-0560-4. The Novick correlations are .39 to .49 for insight and adherence and .38 to .48 for insight and therapeutic alliance, at baseline in 903 European outpatients. It is a post hoc analysis of an observational study designed to compare two oral formulations of olanzapine. Per the competing interests statement, the study was funded by Eli Lilly and Company, four of the six authors are Lilly employees, and the statistical analysis was conducted under a contract with Lilly. The statement then declares no competing interests. The authors’ limitations section notes that the same evaluator assessed all factors, which may have increased the apparent relationship between the variables.
Threads of Counsel 