The Value of Conditional Release for Insanity Acquittees

• forensic psychiatry
• conditional release
• recidivism
• monitoring
• insanity acquittee

The excellent article in this issue by Dr. Haseeb Haroon and colleagues¹ on the outcomes of forensic conditional release in North Carolina is a welcome addition to the literature on this seldom seen topic. As the authors point out, North Carolina is one of a small minority of states that do not use some form of official supervision, usually termed conditional release, to monitor the success of patients who have been discharged from an inpatient not guilty by reason of insanity (NGRI) commitment. Their article provides data on the recidivism rate of North Carolina insanity acquittees on release from inpatient hospitalization, as well as possible indications of racial bias against minority race acquittees in the insanity commitment and release process. The article provides an incentive for improvement not only in those jurisdictions that similarly lack a structured mechanism to follow insanity acquittees after inpatient discharge, but also an opportunity for all states to examine whether their processes in this area are fair and equitable to all persons.

I served as Chair of the Missouri Department of Mental Health (DMH) Forensic Review Committee covering the western half of Missouri for 19 years, where I had the opportunity to publish a similar review of the forensic review program in Missouri.² I have also worked as a staff forensic psychiatrist in Colorado for several years, monitoring NGRI and conditionally released patients.

Haroon et al. refer to findings from the Treatment Advocacy Center³ as suggesting that the relatively high rate of recidivism seen in their study may result from the lack of a formal forensic monitoring system in North Carolina. Having treated, supervised, or testified on behalf of hundreds of conditionally released patients in Missouri and Colorado over nearly three decades, I would emphatically agree. Anecdotal information indicates forensic case monitors in Missouri have intervened countless times before a future crime could be committed, in patients who had either relapsed into substance use, become nonadherent with their prescribed medications, or simply experienced an inexplicable relapse of mental illness symptoms. In most cases, at least in Missouri, the NGRI verdict is only available to those with the most serious diagnoses; such individuals likely would have not recognized their decompensation, or even if they did, may not have been willing to seek help on their own.

Forensic case monitors are knowledgeable of their patients’ warning symptoms, regularly meet with them, and have the statutory authority to supervise their outpatient treatment closely; they have proven invaluable in Missouri for the early detection and safe management of situations that could easily lead to a future offense. Additionally, the mere presence of a firm, well-structured, yet still patient-centered monitoring system has enabled many more such patients to be successful on release for longer periods of time than would have been the case if left to their own devices.

There are several current challenges to providing community psychiatric treatment, including bed space, managed care, and staffing concerns. Additionally, individual patient autonomy is, admirably in most cases, at the forefront in the community treatment setting. These factors often mean that patients experiencing breakthrough symptoms, especially if they lack the insight to seek or cooperate with needed treatment, may simply not be able to access the level of care they need in time to prevent a future conflict with the law. It is also quite possible, as the Haroon et al. article implies through their analysis of North Carolina length-of-stay numbers (“second-longest average LOS in the country”; Ref. 3, p 22), that a professional, well-respected conditional release system may provide the assurance that a fact-finder (usually the committing court judge) may need to consent to a release. As Dr. Haroon’s team points out, the length of stay of an NGRI patient in North Carolina appears strongly associated with the severity of the index crime(s). It is not difficult to imagine how these length-of-stay numbers might well be reduced if North Carolina judges could place their confidence in a strong forensic monitoring system when they are faced with a seemingly risky decision to approve a release, especially when they are placed and retained in office via public elections.

Haroon et al. describe an NGRI commitment system in North Carolina, which in many respects is very similar to that of Missouri and Colorado. Like Missouri and Colorado, there is no statutory limit for the length of hospitalization for an insanity acquittee. Although North Carolina does have statutory mandates for the periodic assessment of a patient’s readiness for release (not provided by law in Missouri or Colorado), the fact is that an NGRI commitment can be indefinite and theoretically lifelong. The burden of proof on the question of release rests with the patient in North Carolina. In states where the burden of proof to be released from an NGRI commitment rests on the patient, as opposed to the government, the forensic mental health system will arguably be in a much stronger position to establish and enforce appropriate conditions on such a release.

Of course, no set of conditions will prevent problems without an effective system in place to monitor adherence and legally viable procedures to address violations. A cadre of knowledgeable, trusted employees serving as monitors is essential. Staff who are employees of the forensic mental health system and dedicated to this role are desirable. If such an important role is assigned as someone’s extra duty, the risk that this duty will be given a lower priority, or possibly neglected, becomes more of a concern. In addition to the monitors who are charged with knowing and meeting with their patients, supervising their care and functioning, and reporting on concerns, an equally important cadre of licensed professional staff with the power to make decisions on this information is also essential. In Missouri, this role is provided by committees responsible for various regions of the state and made up primarily of psychiatrists, psychologists, and social workers. Although a formal history of forensic training and experience is desirable in such a role, this experience is not necessarily a prerequisite. Many good clinicians without formal forensic training will do fine on such a committee, as forensic training, per se, is more often focused on the evaluation of patients regarding legal questions, whereas in the arena of conditional release, the focus is often more on a person’s clinical condition and functioning.

An intriguing concept is proposed in this article, which deserves further exploration. The authors propose a measure they term the “medication stability period” (MSP), which they define as “the period between the discharge date and the most recent preceding psychotropic medication change” (Ref. 1, p 345). For their purposes, they include even minor adjustments of the same drug in calculating this period. For example, a switch from rapid disintegrating tablets to standard oral tablets of the same drug at the same dose would restart the calculation.

This suggestion is an interesting proposal. I have used the same concept, without a formal name, in assessing patients for conditional release. Judges in Missouri often ask, “How long has the patient been ‘stable’ on his medications?” In North Carolina, Dr. Haroon and team found that the average patient successfully gaining a conditional release had gone 1.8 years (SD ± 1.5 years) since the last medication adjustment. This duration is in line with the usual circumstance in western Missouri, in my experience. Of course, there may well be a wide range of average times given unique circumstances, such as a judge approving a release but with the proviso that the patient be switched to injectable or quick dissolving medication before discharge. Such a change coming right before or at discharge could skew this average time considerably.

One recommendation I would make for future researchers using this measure would be to consider redefining the MSP as the period between the last psychotropic medication adjustment and the date the patient is approved by the clinical authorities responsible for the release recommendation. In Missouri, at least, and presumably other states, there can be a considerable amount of time between when a release request to the court is approved for a patient and when that release is heard and granted, and even longer still, in some cases, before a discharge can be effected. In my opinion, this concept will be most clinically useful at the time the clinician is evaluating the patient for readiness for release, which may be weeks, if not months (or longer in some cases), before the release is granted and the patient leaves the hospital. It might seem easier for data analysis to measure this period from a medication order date to a discharge date, but it should not be too hard to measure from the date of a release report, or from a chart note attesting to readiness for release when a formal report is not required.

Haroon and colleagues raise the concern of a potential racial disparity recognized in their data. They term it a “striking” finding, in that for minority race NGRI acquittees the MSP was more than twice as long as that for White acquittees. This finding raises the concern that minority race acquittees must be stable for almost a year longer than their White counterparts to be granted release. In their data, the severity of the index offense, recognized as one potential confounder, did not vary significantly between the two groups. The authors suggest possible reasons for this disparity in the data, including the possibility that courts look for detailed community treatment plans and placement arrangements before approving release, and such arrangements may be more difficult to secure for minority race acquittees. Another possibility they suggest is that minority race acquittees are more likely to have prior criminal histories, and judges, as well as forensic clinicians, might be placing an “overreliance” on historical criminological factors when assessing risk. The authors suggest a potential solution using structured professional judgment tools, such as the Historical Clinical Risk Management instrument (HCR-20V3),⁴ although they go on to suggest that minority race patients may still face individual or systemic biases even when evaluators use seemingly objective behavioral observations. They state, as an example, that recent physical or chemical restraint might be considered as a measure of potential for future violence, but Black patients are more likely to be restrained than White patients.⁵ It would be worthwhile to explore this possible disparity in other states, especially those with robust conditional release programs, to identify structural biases in the NGRI release process.

The primary purpose of the article by Haroon and colleagues, of course, was to research the outcomes of NGRI-released patients in North Carolina, a state which does not have a conditional release program. Their results, albeit with a relatively modest study population of 61 insanity acquittees, indicate that patients in North Carolina released from NGRI commitment recidivate at a rate of 14.8%, which is similar to the rates for unmonitored patients in other states but much higher than for patients under supervision in the community.^2,6 The authors then analyzed reasons that might account for this recidivism rate. Insanity acquittees in North Carolina with problematic substance use had significantly higher rates of recidivism, consistent with prior research elsewhere. Interestingly, however, and inconsistent with some prior research, the authors found no association between recidivism and male sex, serious or violent index offense, psychotic diagnosis, or personality disorder. They also note the fact that, despite an apparent lack of correlation with recidivism rate after discharge, factors such as diagnosis and seriousness of offense are still strongly correlated in their study with longer lengths of hospitalization before discharge. A related finding was that longer periods of hospitalization, and longer periods of clinical stability (i.e., longer MSP), were not associated with reduced recidivism. The authors conclude, in consideration of similar research in other states, that hospitalization for insanity acquittees may be longer than necessary.

The authors recognize that a major limitation in their study was the small sample size. They advise that this limitation could not be helped given the infrequency of insanity acquittals in North Carolina. Undoubtedly, more populous states with court systems more inclined towards insanity acquittal may well provide larger populations to study in these areas. Another limitation, with which I agree, is the possibility that criminal recidivism and rehospitalization in this study were potentially underestimated because of factors beyond their control. They had to rely on reconviction data rather than re-arrest data, which were not available to them. Certainly, the measure of “time to re-arrest” might be substantially shorter in many cases than the “time to reconviction,” given the pace of the criminal justice system. This difference could give a falsely optimistic sense of how long an individual patient did well in the community (a measure of “success” with release) before running afoul of the law. The authors also were unable to access admission data from non–state-operated hospital facilities, or even two of the three state hospitals in North Carolina where released patients might arguably be readmitted. Again, this lack of data could give a false sense of success on release if a readmission was never identified. Finally, because the data were focused on those patients among the 61 who returned to the attention of authorities either by readmission or reconviction, some of those 61 patients may have left the state entirely. If so, their recidivism rate would presumably be unknown to North Carolina authorities. The article is not clear in this regard, and it is likely impossible to know given the lack of an interstate monitoring system.

Despite the unavoidable limitations of this data set, it is my opinion that Haroon et al. have contributed valuable insights to a sparse area of forensic psychiatric literature. I will be eager to see further work from this team and similar research in other states. If other states are like Missouri, forensic databases are kept in a centralized location (as opposed to hospital data for civil psychiatric settings) and should be available for similar analysis.

On a final note, it is important to consider the authors’ discussion of the impact of forensic admissions on the functioning of civil psychiatry in general in their state. The authors point out the dramatic increase in forensic inpatient admissions in North Carolina at the expense of beds available for civil patients. They note that approximately 44 percent of adult state hospital beds in North Carolina are now occupied by forensic patients. In Missouri, virtually all adult state-operated beds are now occupied by forensic, sexually violent predator, or formerly forensic patients. Furthermore, the forensic monitoring system in Missouri, which is arguably quite successful in reducing criminal recidivism among discharged insanity acquittees, is labor intensive and expensive. College-educated employees are required to serve as forensic monitors, licensed professional providers are required to evaluate monitoring feedback on the patients and make decisions, and, in some areas, it has been necessary to establish and subsidize dedicated placement facilities to ensure that suitable destinations are available for patients who gain their conditional release. More research like this excellent article by Drs. Haroon, Wolfe, Feizi, and Barboriak will help justify the resources expended by states to treat and supervise their NGRI acquittees in a humane manner, while helping to ensure the safety of the public-at-large.

• © 2023 American Academy of Psychiatry and the Law