Abstract
This commentary on Salem et al. provides background for their study by reviewing Washington v. Harper and outlining some areas that were not addressed by that decision. It contrasts Harper holdings with those in other U. S. Supreme Court decisions in parallel cases and in United States v. Loughner. It provides cautions about extensions of some holdings in Loughner regarding the use of Harper-type administrative procedures. This article also reviews the methods and findings of Salem et al. and encourages further work. Finally, it raises cautions and voices a call to action concerning potential negative consequences of documenting the effectiveness of administration of involuntary medication in prison.
Well known to both mental health professionals and the public is the dramatic increase in the number of prison inmates in the United States over the past four decades.1 At the same time, the U.S. social safety net, including services to severely ill psychiatric patients, has been shredded by cuts in government budgets. One result, less known to the public, is the diversion of psychotic individuals from the mental health system to jail and prison systems (Fig. 1).
Institutionalization in the United States, per 100,000 adults.
Mental health professionals, on the other hand, know well that a growing number of these inmates need psychiatric services, including antipsychotic medications. Because of their psychoses, many of these prisoners lack the insight, judgment, and decision-making capacities necessary to allow them to agree to the use of these medications. Many are incompetent to give informed consent. When such inmates are a danger to others, self, or property or are gravely disabled, it leaves correctional institutions in a challenging situation, as other provisions, such as civil commitment, may not apply or be available. The U.S. Supreme Court in 1990 addressed this need in the case of Washington v. Harper.2 That decision provided a constitutional avenue for the involuntary provision of antipsychotic medications to such mentally ill prison inmates via administrative review. Yet, until now, 25 years after the Harper ruling was handed down, there has not been a published study of how well the involuntary administration of antipsychotic medications works to improve functioning and to reduce violence in mentally ill prison populations.
Salem et al.3 deserve commendation and thanks for their study. Their work is unique in being the first published study of the effectiveness of administering involuntary antipsychotic medications to mentally ill prisoners, according to the procedure described in Harper.
Washington v. Harper
Walter Harper was confined to the Washington State prison system following a robbery conviction. He had a history of bipolar disorder that had responded positively to antipsychotic medications. However, he became violent when not taking them. He was forcibly medicated with antipsychotics according to the procedure of the state prison system, which involved approval by a majority of a tribunal consisting of a nontreating psychiatrist, a psychologist, and the prison warden. The panel psychiatrist had to be part of the majority for involuntary medication to be used.
Mr. Harper sued in state court, contending that the Constitution's Fourteenth Amendment Due Process Clause entitled him to a full judicial hearing, rather than a prison committee decision. The court ruled against him, but he appealed that decision. The Washington State Supreme Court ruled that he did have a right to a court hearing. However, on appeal, the U.S. Supreme Court held that the original state prison policy provided adequate protection of Mr. Harper's liberty interests, balanced against “the State's legitimate interest in combating the danger posed by a violent, mentally ill inmate (Ref. 2, p 211).” The Court found that Mr. Harper's interests were “better served, by allowing the decision to medicate to be made by…medical professionals rather than a judge (Ref. 2, p 231).” In reaching this decision, the Court may be seen as presaging the findings of Salem et al., stating, “There is little dispute in the psychiatric profession that the proper use of the drugs is an effective means of treating and controlling a mental illness likely to cause violent behavior (Ref. 2, p 211).”
Harper versus Other Court Decisions
The Harper decision ratified the adequacy of administrative panel decisions to allow administration of antipsychotics to nonconsenting prisoners, but it left several related areas undefined. It did not define mental illness, leaving open the question of whether, for example, personality disorders might qualify an individual for involuntary medication under certain circumstances. It did not rule out the use of court hearings to provide for such decisions. It did not specify the constitutionally acceptable lengths of time for which prisoners may be involuntarily medicated after an administrative hearing is held. It applied to an adjudicated prisoner, but did not discuss the liberty rights of pretrial detainees with mental illness in similar, psychotic straits. It also did not address whether the proper venue for medicating inmates is a psychiatric facility on the one hand or a cell in a lockup on the other. Finally, it did not deal with involuntary medication for competency restoration of persons found incompetent to stand trial.
The more recent case of United States v. Loughner4 illustrates several of the above ambiguities left by the Harper ruling. Jared Lee Loughner ultimately entered an agreement to plead guilty to the January 8, 2011, killings of six persons in Tucson, AZ, and the wounding of 12 others, including former Congresswoman Gabrielle Giffords. He was sentenced in November 2012 to life imprisonment without parole. However, during almost the entire intervening time of nearly two years, he was given antipsychotic medication involuntarily. After he refused to take the medication voluntarily, a series of Harper-type administrative hearings, rather than judicial actions, were held to commit him to receive involuntary medication on the basis of mental illness (i.e., schizophrenia) and of dangerousness to himself. However, although he involuntarily received antipsychotic medication solely as the result of Harper-type hearings based on his dangerousness, he was simultaneously a pretrial detainee, an inmate in a nonmedical correctional facility, and a person held to be incompetent to stand trial. Still, he was not determined by lower federal courts or by the Ninth Circuit Court of Appeals to have been entitled to a formal judicial review of the administration of involuntary medication, regardless of the drug's impact on his competency.
Loughner appears to conflict with previous rulings by the U.S. Supreme Court in Riggins v. Nevada5 and Sell v. United States,6 which established a higher standard, a right of judicial review for the forced medication of pretrial detainees. However, in a Sell footnote, the court commented favorably on the potential use of Harper proceedings or civil commitment as alternatives for securing involuntary antipsychotics for pretrial detainees. Loughner appears to support the use of Harper in this circumstance. In addition, the Court held in Vitek v. Jones7 that a similar standard of judicial review must be observed in the transfer of prisoners to mental hospitals, whereas Mr. Loughner, an incompetent pretrial detainee presumably entitled to greater Fourteenth Amendment due process protection, was not provided judicial review of the venue in which the involuntarily received antipsychotics would be administered. Such decisions seem to pave the way for correctional settings to provide involuntary treatment in a variety of circumstances and venues. However, although the courts may look favorably on such practices, the question remains whether they are clinically and ethically reasonable. These concerns deserve both consideration and scrutiny.
Extending Harper?
In previous issues of the Journal, Felthous has provided elegant elaborations on these apparently conflicting decisions8,9 that will not be repeated here. He has, however, raised important questions about the post-Loughner potential for extending Harper procedures to allow involuntary antipsychotic medications to be given in jails to pretrial detainees and to incompetent detainees. As the correctional systems in the United States face more and more mentally ill persons in their populations, pressure grows to make psychiatric care available in correctional settings that have widely varying capacities for providing medical services. Political and economic forces, and now, after Loughner, legal forces as well, may converge to move psychiatrists toward providing ever more involuntary antipsychotics outside secure psychiatric settings. For instance, it is conceivable that Harper-type procedures will be applied to arrestees who have yet to be charged.
Although some jurisdictions have laws that would bar the practice, Harper-type procedures could also be applied to the use of other types of psychotropic medications (though most intramuscular psychotropic preparations are in the antipsychotic category). Perhaps paradoxically, using Harper in this way would force courts to re-examine the perception that antipsychotics, as distinct from other types of medications, must be treated uniquely because of their purported personality-altering, or even will-altering and person-altering effects, with the high potential for causing debilitating side effects such as tardive dyskinesia.5
It is critical to attend to Harper and its alternatives to determine the most legally, ethically, and medically appropriate ways of serving the psychiatric needs of mentally ill individuals in custody. It is appropriate for psychiatrists to consider whether they can provide medically and ethically appropriate treatment with antipsychotics in any carceral setting, or whether the safety of detainees and custodial personnel involved would be better served by the use of involuntary civil commitment and transfer to better-equipped, medical facilities for acute treatment.
It is foreseeable that, after stabilization with antipsychotic and other treatments in medical facilities, some detainees would be discharged to their original jail while still under court order. A portion of them, having become more cooperative with treatment, could have their medications involuntarily given orally in jail on an ongoing basis in a humane and safe fashion under assisted outpatient treatment commitment procedures. It is appropriate as well to consider the holdings in cases such as Vitek v. Jones7 and Baxstrom v. Herold,10 in which prisoners retain some due process rights regarding medical decision-making and self-determination. These cases demonstrate that there are constitutional limits to what can be done in correctional settings, but these limits are not yet well demarcated, especially with regard to pretrial detainees and arrestees.
Methods
Salem et al.3 hypothesized that providing involuntary antipsychotics in a prison setting would reduce the number of prison inpatient days and the number of inmates who incur disciplinary charges. Although their results supported their second hypothesis, they reported no significant reduction in number of inpatient days after inmates were given antipsychotic medications involuntarily, compared with the number of inpatient days recorded before the medications were provided.
They chose to use the number of inpatient days before and after administration of involuntary medication as a proxy for estimating inmates' functioning. Likewise, they used the incidence of rule infractions before and after administration as a proxy for estimating inmates' violent behavior. The inmates served as their own before-and-after controls.
The authors were constrained by several circumstances. Their study was unfunded and their data retrospective. They were unable to use a control group because their state administrative code prohibits experimenting on inmates. Assuming that the raw data available to them in such circumstances were narrow in scope, their choice of proxies, cited above, was of necessity quite restricted. It appears that they did very well under relatively limiting circumstances.
Nevertheless, there are aspects of their chosen method that may have led to further limitations in the generalizability of the data they gathered. First, they elected to include in their study only those inmates who were treated involuntarily with medications for a whole year following their Harper tribunal. In selecting a group needing such long-term administration of involuntary medication, they may have selected the most severely ill portion of the mentally ill inmate population from which to gather data. If this were the case, then this group of inmates may have been the least likely to have shown significant benefit from treatment. The positive effects of involuntary administration of medication in a less restricted or more medication-responsive population, such as those treated involuntarily for a shorter time, may therefore have been underestimated.
Because they could not include a placebo control group in their study, they employed a before-and-after-Harper comparison of inpatient days and infractions to determine whether involuntarily received medication alters outcomes. This comparison did not take into account the potentially confounding effects of the increased frequency of infractions or of psychiatric destabilization that sometimes occurs when inmates first arrive in prison. Many new arrivals simply “don't know the territory and don't know the ropes.” Prison entry, the time of greatest adjustment to a new, stressful environment, tends to be the time when inmates, particularly those who have or are prone to mental illness, are most likely to incur infractions. Including this adjustment period in the data gathered by Salem et al. may have heightened the number of infractions committed in the pre-Harper period. Including the period immediately after prison entry may therefore make the effects of involuntary medications appear greater than they might have been were the adjustment period not included in the data.
Another potential confounder in their data came from what the authors termed weighting. Some inmates who were treated involuntarily for a year had not resided in the institution for an entire year before their Harper hearing. The average number of prison days before involuntarily receiving medications was 300, and a large number had considerably fewer prison days. Yet this period was being compared with a 365-day period post-Harper hearing. Because of this, the authors prorated for those relative short-timers the number of inpatient days spent and the number of infractions they committed before their Harper hearings by using a multiplier averaging 1.22. This allowed a statistical approximation of equalization of time spent before and after involuntarily received medications. This act of prorating could further magnify the effects of the adjustment period cited in the previous paragraph, again inflating the differences between inpatient days and infractions in favor of reductions after Harper hearings.
In addition, infractions were relatively low-frequency events in the study population. The weighted mean number of charges per inmate was only 2.1 before the Harper hearings and .82 after inmates were medicated involuntarily. Because the base rate was quite low, the smaller data sample in the pretreatment control period increases the potential error rate during this period. There is no inherent reason to believe that this would introduce a systematic bias, but it does introduce greater uncertainty into the statistical calculations.
Salem et al. describe very well the limitations of their study. In addition to the lack of randomization and placebo control, they mention the potential for selective infracting of mentally ill prisoners and the weighting of data collected in the period before involuntary medication was given. They also cite the lack of attention to variables such as movement between units, the presence of personality disorders, especially antisocial personality disorder, and the reason for placement on involuntary medication. Another limitation they did not mention was the lack of attention to the potential of substance use disorders to confound the results. One might anticipate that active or even past use of psychotogenic substances would decrease the likelihood of response to antipsychotic treatment11 and allow for more meaningful separation of inmate groups.
Potential Consequences of Success
It is possible that future studies demonstrating a reduction in inpatient days after involuntarily administered antipsychotics will ultimately have a negative effect on the psychiatric care of prisoners. The authors cited an unstated practice in the prison system they studied that might well account for the lack of change in the number of inpatient days before and after involuntary medication. When inmates are placed on an involuntary medication protocol, the tendency in the system studied, the New Jersey state prisons, is to keep them on an inpatient unit during the entire duration of that protocol, regardless of whether there is behavioral improvement. There are many reasons for this practice, including the increased attention and therapeutic programming that these very ill inmates receive in addition to medications, the security needs of other inmates and correctional personnel, and the relative safety of the units housing the mentally ill. For those requiring injection, such a setting may be preferable for medical and hygienic safety reasons, as well.
Salem et al. point out that these conditions do not parallel those in the general community, and that makes the results less generalizable. Indeed, numerous community studies regarding assisted outpatient treatment (AOT) that were cited by the authors showed that involuntarily administered antipsychotic medications are effective in reducing violence and hospital days.12,13 One could speculate that, had a similar study been conducted on an inpatient unit in a prison operated on a private, for-profit basis, there might have been significantly fewer inpatient days following the involuntary administration of antipsychotic medications.
The practice of keeping involuntary patients on prison hospital wards for the duration of their involuntary administrative order appears to be a humane and justifiable practice. However, it commits inpatient resources to those on involuntary antipsychotics regardless of the acuity of their condition, rather than focusing the limited psychiatric treatment resources available on prison wards on the sickest and neediest.
In the future, we expect that the effectiveness of involuntarily administered antipsychotic medication will be more convincingly demonstrated. If that occurs, one might anticipate that the external economic pressures that exist in community hospitals will be brought to bear on prison inpatient psychiatric units as well.
As the number of U.S. prison beds has expanded over the past several decades, prison budgets have grown in parallel. Budgetary items that benefit the health of prisoners have not been subjects of popular political support. Despite improving economic conditions during the past several years, most states' governors and legislators can be expected to train their budgetary knives on prison psychiatric units if they have reason to believe that they contain “pork.” Such budget reductions may come to pass if follow-up studies of the work of Salem et al. demonstrate the success of involuntarily administered antipsychotic medications and may also drive the increased use of Harper. This possibility can be seen as parallel to the role of the successful introduction of antidepressants, lithium, and neuroleptics in prompting the deinstitutionalization movement of the 1960s and 1970s. It is generally accepted that the money saved by deinstitutionalization was not used to fund community psychiatric care. This failure contributed to the unraveling of the social safety net for psychiatric patients in the United States.
If such pressures for earlier discharge of involuntarily medicated inmates from prison psychiatric units result in cuts in the funding of inpatient units, it will be critical for correctional psychiatrists and other mental health professionals to work to ensure that financial savings that result from shortened inpatient stays be used to enhance the safe, secure, and humane administration of involuntary antipsychotic medications in general prison populations. The use of the funds might include the institution of step-down units; the increased use of general-population psychiatric case managers; and the enhanced availability of psychiatrists, nursing, and therapy staff to psychiatrically ill prisoners in the general population. If involuntarily administered injectable antipsychotic medications see increased use in general prison populations as a result of deinstitutionalization and management of care, resources to allow consistently hygienic settings for intramuscular injection and appropriate clinical monitoring should also be enhanced.
The effects of stigmatization of prisoners branded as mentally ill also should not be ignored. Inmates are already severely stigmatized by having resided on inpatient psychiatric units. Although involuntarily treated inmates discharged from inpatient psychiatric units may re-enter the general population when they behave more normally, coming from psychiatric units or receiving involuntary injections makes them potential targets in the general population. They still may refuse or “cheek” oral medications and may require short- or long-acting injectable antipsychotics. Observation by other inmates of prisoners receiving injections in their gluteal muscles in their cells would tend to greatly heighten that stigmatization. It is hoped that efforts can be made to convince custodial personnel and prisoners to arrange such procedures in a nurse's station rather than in prisoners' cells.
Future Directions
With their pioneering and much needed work, Salem et al. have paved the way for further research in an important and thus far unexplored area of correctional psychiatry. Despite its limitations, which the authors well acknowledge, this is an important study that has great value for the field. We hope that prospective studies of the effectiveness of involuntarily administered antipsychotic medications can be undertaken that allow for the collection of data that more directly measure involuntarily medicated prisoners' level of psychosocial functioning and violence. Harper mentions grave disability as an alternative finding to violence that may justify the administration of antipsychotics to mentally ill prisoners. It also mentions balancing the legal due process rights of prisoners against the penological interests of the prison. This includes maintaining order and avoiding damage to prison property. Collecting and analyzing data in such areas as involuntarily medicated prisoners' participation in educational and work programs, changes in custody level, incidence of early release, and even forced showers may provide detailed insight into the effects of involuntarily administered antipsychotic medications. The impacts of co-occurring disorders such as antisocial personality disorder, other personality disorders, and substance use disorders on the effectiveness of involuntarily administered antipsychotic medications is another fertile area for future research in this population.
Finally, the literature on AOT supports the premise that the effectiveness of involuntary medication is enhanced by additional programming that addresses the pathologies of mental illnesses with psychosocial interventions.14 In future publications on the effectiveness of involuntary treatment in prisons, attempts should be made as well to shed light on how combinations of treatments work to optimize outcomes. We encourage Dr. Salem and her coauthors, as well as others, to take part in this ongoing work.
Footnotes
Disclosures of financial or other potential conflicts of interest: None.
- © 2015 American Academy of Psychiatry and the Law
