Mental health courts (MHCs) are appearing with increasing frequency across the United States (Redlich, Steadman, Robbins, Monahan, & Petrila, 2006). The number of courts has grown from two in 1997 to approximately 150 in 2008 (Council of State Governments, 2008; National GAINS Center, 2008). MHCs are specialty criminal courts for persons with mental illness who have been arrested (Griffin, Steadman, & Petrila, 2002) that fall under the rubrics of problem-solving courts and diversion programs. Problem-solving courts, such as Drug Treatment Courts, Domestic Violence Courts, and Homelessness Courts, are courts that focus on a category of offenders or criminal charges (thereby becoming better equipped to handle common issues that arise across cases), and that tend to take a therapeutic approach (see Odegaard, 2007; Winick & Wexler, 2003). These courts are often considered a new way of doing justice; courts that stand apart from traditional adversarial approaches. Diversion programs are formalized efforts often for persons with serious mental illness (but not exclusive to them) that attempt to identify and enroll eligible individuals into outpatient treatment programs (Petrila & Redlich, 2008; Redlich, 2007). Diversion is from the criminal justice system (either from being charged or being incarcerated) into community mental health and substance use treatment.

Mental health courts tend to be diverse, with varying requirements and procedures (Erickson, Campbell, & Lamberti, 2006; Wolff & Pogorzelski, 2005). However, there is one mainstay across all the courts: MHCs are intended to be voluntary (Redlich, 2005). Persons referred and deemed eligible theoretically have the choice to take their case to the MHC or to keep their case in regular criminal court processing. The purpose of this study was to determine whether newly enrolled clients of MHCs were aware they had a choice, and whether their enrollment was knowing and intelligent, as well as voluntary. In addition, because the decision to enroll in a MHC is a legal decision that is often accompanied by a mandatory guilty plea, a further purpose was to examine levels of legal competence in newly enrolled clients of MHCs. As the Supreme Court decided in Godinez v. Moran (1993), competence to stand trial is an adequate measure of competence to plead guilty. Finally, we were interested in determining the demographic, legal, and clinical factors that influence comprehension in MHC enrollment decisions.

Mental Health Courts: Voluntary

Why is there a requirement that entering into a MHC be voluntary? For some, this is an odd question. MHCs appear to be in the best interests of the client as the person gets to avoid jail and get treatment services. As noted by the Bazelon Center for Mental Health Law (2003), an advocacy center for people with mental illness,”On its face, a defendant’s selection of a therapeutic court over one structured around determining guilt and meting out punishment would appear an obvious choice” (p. 5). They go on to state, “Further complicating the voluntary election of mental health court involvement is the fact that such decisions are made when the defendant is likely to be under considerable stress, having been arrested and taken into custody, and perhaps having spent some time in a jail cell, often without treatment of any kind” (p. 5) (see also Stefan & Winick, 2005).

Many MHCs require new clients to sign written contracts upon enrollment. MHCs represent a commitment to take medications, to attend and engage in mental health/substance abuse treatment appointments, to return to the court for status review hearings, to meet with a case manager and/or probation officer, to show up to court on time, and to follow other individualized treatment orders. It is also a commitment with consequences. Although MHCs are generally flexible and understanding of the populations they deal with, there are sanctions available for non-compliance, including increased supervision and re-incarceration. The willingness to use jail as a sanction varies. Redlich et al. (2006) surveyed the U.S. population of MHCs and found that courts that had more frequent judicial supervision and enrolled more defendants charged with felonies were more apt to use jail as a sanction.

From a legal standpoint, entry into the courts must be voluntary; if they were not and all offenders with mental illness were required to partake in a MHC, the equal protection guarantee of the 14th amendment would be violated in that a certain subgroup of offenders would be singled out and treated specially (see Seltzer, 2005). Further, the majority of MHCs (67%) require guilty pleas as a condition of enrollment, with an additional 16% requiring guilty pleas for some of their clients (Council of State Governments, 2006). Some courts expunge guilty pleas upon successful completion, however. Decisions that pit release from detention and access to treatment against pleading guilty might be defined as coercive to some.

Although MHCs are intended to be voluntary, little is known about this aspect of the courts. With therapeutic jurisprudence at their base, MHCs have the potential to become paternalistic in nature (see generally Petrila, 1996; Stefan & Winick, 2005). At its extreme, an overly paternalistic court could either not present the choices or override the wishes of a potential client deemed to be too unstable to make the ‘proper’ decision to enroll. Although there is mixed evidence, it appears that when given the option of MHC, few decline. In a study of seven mental health courts, only one of 148 accepted persons (less than 1%) chose not to enroll (Steadman, Redlich, Griffin, Petrila, & Monahan, 2005). In contrast, in studying two Seattle-based MHCs, Trupin and Richards (2003) found that more individuals opted out than opted into the courts. Of particular relevance to this study, Trupin, Richards, Lucenko, and Wood (2000) noted that those who opted in were more insightful about their mental health problems.

Researchers (Boothroyd, Poythress, McGaha, & Petrila, 2003; Poythress, Petrila, McGaha, & Boothroyd, 2002) examined enrollment decisions and perceptions of coercion in the Broward County, Florida MHC. Several findings are relevant. First, the authors collected and coded court transcripts of initial hearings. Boothroyd et al. (2003) reported that the primary purpose and focus of the court was explicitly mentioned in only 28% of transcripts, explicit statements of voluntariness in 16% of transcripts, and any mention of competence-to-proceed in 29% of transcripts. Second, they obtained self-reported perceptions of MHC voluntariness, and found that approximately 75% of Broward County clients either reported not being told that court participation was voluntary or reported being told only after their first hearing. Based on these findings, in this study, we anticipated that the majority of MHC participants (i.e., 50% or more) will claim not to be aware that the decision to enroll in the MHC was voluntary.

Lastly, Poythress et al. (2002) measured perceived coercion in MHC-entry decision-making in the Broward County MHC, which includes constructs such as perceived freedom, perceived influence, and perceived choice. Overall, the average perceived coercion scores were quite low, M = 0.69 (SD = 1.30) on a scale of 0 (low coercion) to 5 (high coercion). However, clients of MHCs’ claims of awareness of the voluntary nature of the court affected the degree of perceived coercion: clients who claimed not to know they had a choice in enrolling had significantly higher perceived coercion scores than those claiming to be aware. These results remained even after controlling for severity of mental illness. We expected similar results in this study.

In summary, although mental health courts are intended to be voluntary, there is scant information available as to whether they actually are. The information that has been collected indicates that court participants often either claim not to have been told that the decision was theirs to make or claim to have been told only after they were involved. Examining whether the decision is made knowingly and intelligently is imperative for at least two related reasons: 1) the target population (i.e., persons with serious mental illness) is known to have deficits in legal comprehension, and 2) the MHC entry decision may be made under significant stress and instability.

Mental Health Courts: Knowing and Intelligent

Two considerations are important to making a knowing and intelligent decision to enter a MHC. First is general legal knowledge, most notably, adjudicative competence. The standard for competence is the Dusky standard, which mandates that defendants have a rational and factual understanding of the proceedings against them, as well as the ability to consult with their attorney (Dusky v. U.S., 1960; see also Godinez v. Moran, 1993). As stated, most MHCs require guilty pleas, especially from felony enrollees (Redlich, Steadman, Monahan, Petrila, & Griffin, 2005; Redlich et al., 2006). In essence, the decision to plead guilty and the decision to enter the court can become one and the same. The second consideration is specific knowledge relating to the MHC itself. An informed decision would entail knowledge of MHC procedures, requirements, confidentiality releases, consequences for compliance and non-compliance, alternatives to participation, and what happens upon graduation and termination. As stated by Susan Stefan (Stefan & Winick, 2005), “The people don’t go into the process understanding what mental health court is all about, and no one explains it to them in terms of benefits and drawbacks…. This is not an atmosphere that is conducive to knowing and intelligent decision making” (p. 516).

To our knowledge, the adjudicative competence of current clients of mental health courts has not been examined. Competence is a threshold issue. All defendants are presumed competent unless the question is raised. The leading reason to raise questions of competence is for psychiatric reasons (Pinals, 2005). Stafford and Wygant (2005) examined competence among a subset of persons referred to the Akron, Ohio MHC (i.e., potential clients of MHCs). Over a three-year period, a total of 472 persons had been referred to the court; of these, 85 (18%) had their competence to stand trial raised. And of these 18% who had their competence raised, more than 75% were found incompetent to proceed. Interestingly, only four defendants whose competency had been raised ended up in the MHC, two who had been found competent, and two who had been found incompetent and then restored.

Stafford and Wygant (2005) also examined the factors that associated with findings of incompetence. Like other studies on the competencies of persons with serious mental illness (Grisso & Appelbaum, 1995; Hoge et al., 1997; Nicholson & Kugler, 1991; Rosenfeld & Wall, 1998), diagnosis was influential. Having a psychotic disorder, which included schizo-spectrum disorders, strongly distinguished among persons found competent versus incompetent. Other factors found to influence competence are socioeconomic status (or employment) and charge level, such that persons who are unemployed or from lower SES, and persons with minor charges were more likely to have deficiencies in competence (Cooper & Zapf, 2003; Hoge et al., 1997). Gender has not been found to influence competence (Poythress et al., 1998).

To our knowledge, MHC comprehension has not yet been examined, with the exception of knowing whether the court is voluntary or not. In contrast, there is a modest literature on ability to provide informed consent in health-related contexts, including consent for treatment among persons with mental impairment. Traditionally, informed consent law includes three core constructs. First is the provision of sufficient information to make a decision. As stated by Appelbaum and Grisso (1995), “(t)he amount and precise details of the information to be disclosed are determined by what a reasonable person would find material to a decision about treatment.” (p. 106). Second, informed consent involves a voluntary decision, one that is absent of coercion. And, third, persons should be competent to make treatment decisions. In an extensive examination of competence to make treatment decisions, Grisso and Appelbaum (1995) found that only 48% of the schizophrenic sample performed adequately, compared to 76% of the depression group, 88% of the medically ill group, and 96% of the non-ill group. In addition, as found with adjudicative competence, they found less severe symptoms and higher SES to be associated with fewer competence deficits.

In addition to these three core constructs of informed consent, Petrila (2003) argued that a truly informed consent often involves aspects over and above those directly pertinent to the treatment itself, such as possible financial conflicts of providers or alternatives to treatment. Similarly, persons deciding to enter a MHC should have available to them information about the treatment itself as well as information about treatment providers’ relationship with the court and with community supervisors (e.g., probation officers), criminal justice consequences (e.g., reduction of charges, sanctions), and other areas tangential to treatment. Simply put: When enrolling in MHCs, there is much to grasp, including (but not limited to) an understanding that attending court and community treatment appointments, and taking prescribed medications are mandatory, that non-compliance results in sanctions and increased supervision, and that sentences and/or charges could be reinstated.

Redlich (2005) surmised that adjudicative competence, competence to make treatment decisions, and MHC comprehension would overlap, as they share some of the same basic constructs. Based on the findings in the adjudicative competence and informed consent literatures, we expected that 1) some, but not all, MHC participants would not fully understand and appreciate MHC procedures and requirements (i.e., MHC knowledge would vary on a continuum ranging from no to full understanding; see Redlich, 2005), and 2) level of MHC knowledge would be influenced by personal characteristics, including diagnosis, employment status, and charge level.

Study Overview

The primary purpose of this study was to gain an understanding of whether and to what degree clients of MHCs comprehend and appreciate the voluntary nature, requirements, and procedures of the courts at the onset of their participation. We were also interested in examining the factors that affect differential levels of understanding. We surveyed 200 clients of MHCs from two participating courts, most within one month of entering the court. We measured awareness of the voluntary nature of the court and comprehension of mental health court practices and consequences. We also assessed adjudicative competence to examine in relation to MHC voluntary awareness, understanding and appreciation.

Methods

Study Participants and Mental Health Courts

Participants were newly enrolled clients from the Brooklyn, NY MHC and Washoe County (Reno-Sparks area), NV MHC. Both courts require clients to plead guilty in order to enroll. Table 1 presents information on all clients who entered the MHC during our data collection. At both sites, clients of MHC who were interviewed did not differ significantly from those who were not interviewed in mean age, gender, or arrest charge severity. Table 2 lists demographic, charge, and clinical characteristics of those interviewed in the study.

Table 1 Status of clients who entered MHC during study timeframe by site
Full size table
Table 2 Interview participant characteristics by study site
Full size table

The Brooklyn MHC was established in 2002 (Fisler, 2005; O’Keefe, 2006). The court utilizes a six-page document to describe the court and its requirements to new clients. This document includes headings such as “What do I have to do?” “What are the rules of the Mental Health Court?” and “How long will I be involved in the Mental Health Court?” Advantages to enrollment were specifically noted in a section entitled, “What’s in it for me?” which included access to services, recognition of progress, dismissal or reduction of charges, and opportunity to move forward with their lives. Disadvantages were less specific, but sanctions for non-compliance and consequences of failure (i.e., return to jail or prison) were explicitly mentioned.

The Brooklyn MHC also requires new enrollees to sign an “Ongoing Release of Confidential Information.” Offenders charged with misdemeanors, a felony for the first-time, and those with more than one prior felony conviction must agree to treatment mandates of 12, 12–18, and 18–24 months, respectively (O’Keefe, 2006).

The Washoe County, NV MHC was established in 2001. Accepted clients of MHCs are provided a one-page information sheet that includes a one-paragraph description of the court. The sheet also informs when and where the court is held, and that the program is a minimum of one year. After April 25, 2006, new enrollees were also asked the sign a “Mental Health Court Agreement” that lists 15 stipulations, including “I will sign any releases of information as required in order for the court to obtain information needed for my participation,” “I will take medications for my psychiatric condition as prescribed by a doctor,” and “I understand that should I fail to comply with these conditions, I will be subject to sanctions, including jail, community service, or any other sanction the court deems appropriate.”Footnote 1 Advantages and disadvantages to enrolling in the court were not explicitly stated, though sanctions for non-compliance were mentioned.

Data Collection Measures

The interview instrument contained sections assessing demographics (e.g., age, race/ethnicity), criminal justice history (e.g., number of prior arrests), insight into mental illness (Insight and Treatment Attitudes questionnaire, ITAQ; McEvoy et al., 1989), and current symptomatology (Colorado Symptoms Index, CSI; Conrad et al., 2001). The ITAQ measures recognition of mental disorder, as well as attitudes toward medication and general treatment compliance, and is strongly correlated with other measures of insight (see Cuesta, Perlata, & Zarzuela, 2000). The CSI is an often-used measure of symptoms that has good internal consistency (α = .87), is stable over time, and relates to family and provider ratings and use of services (Lee, Shern, & Coen, 1999).

The sections of particular relevance to this article were the MacArthur Competence Assessment Tool—Criminal Adjudication (MacCAT-CA); Awareness of Choice in MHC enrollment (A-MHC); and Comprehension in MHC-Decision-Making (C-MHC). The MacCAT-CA is a normed and standardized scale (Otto et al., 1998), whereas the A-MHC and the C-MHC were developed for this study. There were two versions of the interview instrument: Counterbalance (CB) 1 and 2. In CB1, the MacCAT-CA directly preceded the A-MHC and C-MHC, whereas in CB2, the A-MHC and C-MHC preceded the MacCAT-CA.

MacArthur Competence Assessment Tool—Criminal Adjudication (MacCAT-CA; Poythress et al., 1999)

The MacCAT-CA tests 1) Understanding, 2) Reasoning, and 3) Appreciation as related to the Dusky standard of adjudicative competence (see Bonnie, 1992). The MacCAT-CA was developed for use with adult defendants with and without known mental health problems. In the beginning of the instrument, respondents hear a brief story about two men who get into a fight at a bar. The Understanding and Reasoning questions follow from this story. Appreciation questions relate to the defendant’s own pending criminal case. Because both data collection sites require a guilty plea from defendants, these questions were not appropriate and were not included. The authors of the MacCAT-CA stress that the measure should be used as a tool and not as a test of competence. Thus, we point out that we were not measuring legal competence per se, but rather measuring the components behind the construct.

As described in Otto et al. (1998), the MacCAT-CA has well-established psychometric properties. Understanding and Reasoning have strong internal consistency, with Cronbach’s alphas of .85 and .81, respectively and interrater reliability correlations ranging from very good to excellent (i.e., from .85 to .90). The MacCAT-CA correlates significantly with past measures of adjudicative competence and with clinical measures.

Awareness of Choice in MHC Enrollment (A-MHC)

The A-MHC consisted of a series of questions aimed at determining whether the decision to enroll in the court was voluntary. To be able to compare with previous studies, we first asked a question similar to that used by Poythress et al. (2002): “Did anyone ever explain to you that you could choose to have your case in the mental health court or that you could have your case kept in regular criminal court?” If participants answered ‘yes,’ they were asked when this explanation was given (before, at, or after enrollment). We also asked a similar question of whether and when MHC requirements were explained.

We included the MacArthur Perceived Coercion Scale (MPCS), adapted for MHCs, which was also used by Poythress et al. (2002). This scale includes five statements concerning the choice to enroll in MHCs (e.g., I chose to be in the mental health court; I had a lot of control over whether I went to the mental health court). A 5-point Likert scale was used with 0 = strongly agree to 4 = strongly disagree. Although the label of the scale is perceived coercion, the inverse is perceived voluntariness. Since its inception, the MPCS has been used with thousands of patients and in numerous countries. It has been found to be sensitive and internally consistent (e.g., Cronbach’s α = .90; Gardner et al., 1993) and to have obtainable interrater reliability (kappas range from .73 to 1.00; Lidz et al., 1997). In this study, the Cronbach α for the scale was .82.

Finally, participants were asked in an open-ended format, “What would have happened with your legal case if you were NOT in the mental health court?” (and then immediately asked again, “Is there anything else that could have happened…?”), and “After the court has said that you are eligible to participate in the mental health court, who makes the final decision about whether or not you participate?”

Comprehension in Mental Health Court-Decision Making (C-MHC)

The C-MHC included open-ended questions, yes/no questions, and three series of true/false/I don’t know questions aimed at assessing participants’ accuracy in understanding of MHC procedures, requirements, and consequences. Open-ended questions asked about advantages and disadvantages of being in the MHC over regular criminal court (participants were given two chances for each), and what should people do if they do not want to be in a MHC any longer.

Yes/no questions included 1) Can any person charged with a crime be in a mental health court if they want to be? 2) Was it your choice to be in the mental health court? and 3) Can a person stop being in a mental health court if he or she wants to?

The three series of true/false/I don’t know statements included these opening stems: 1) To be a mental health court participant, people are agreeing to… (14 statements); 2) If people do not follow the conditions of mental health court, they can… (12 statements); and 3) If people follow the conditions of mental health court, they can… (7 statements). The statements consisted of actual MHC procedures, requirements, and consequences (e.g., returning for status review hearings, having to take prescribed medications and go to treatment appointments, pleading guilty as a condition of enrollment, and having their original charges or convictions dropped). Statements of non-MHC procedures, requirements, and consequences were also included (i.e., false statements), such as having to take lie detector tests, never being arrested again, and being charged again for the same crime.

We created one MHC Knowledge summary score by summing the 33 correct–incorrect scores. The Cronbach’s α for this summary score was .62. When the question “If people do not follow the conditions of mental health court, they can be sent back to criminal court” was excluded, the Cronbach’s α rose to .66. Thus, we deleted this question from the MHC Knowledge score, which in hindsight was ambiguous and confusing. Because participants in both courts had pleaded guilty, they would not be sent back to court, but rather to jail or prison.

Non-Self-Report Measures

The MHCs provided the following information on study participants: 1) MHC Entry date (official enrollment date); 2) Diagnoses; 3) Global Assessment Functioning (GAF; Axis V, DSM-IV) scores, and the dates GAFs were assessed; 4) Most serious target arrest charges and misdemeanor/felony status; and 5) Target arrest date. In addition, we obtained from the courts the number of times participants came before the Judge of the MHC prior to our interview.

Procedures

All study procedures and instrumentation were approved by human subject review boards. Data collection ran from December 2005 to March 2007. We attempted to approach all newly enrolled (excepting those known to be aged 17 years or younger, or not to speak and/or understand English) clients of MHCs within that timeframe. For the most part (92%), interviews were conducted within one month of the official MHC start date.Footnote 2 We did not want to let too much time elapse in between enrollment and the interview, and thus tried to interview within one month of entry. At the same time, we knew it was not feasible to interview all participants on their enrollment date given constraints beyond our control (e.g., transportation and custody issues). We were able to interview 27% of the sample on the same day s/he entered the court. The mean number of days between entry and interview was 12.32 (SD = 16.91), which did not differ by site (the median was 7 days and the mode was 0 day).

Written informed consent was obtained from those participants willing to take part in the study. To ensure that participation was truly voluntary and informed, a four-question ‘quiz’ was included after the consent form was reviewed, but prior to signing the consent. If participants got any one question incorrect after three attempts to explain, s/he was considered ineligible to be in the study; no one was considered ineligible for this reason. Next, researchers conducted the in-person interviews. On average, interviews lasted 66 min (SD = 16). Participants were compensated $20 for their time.

Results

Preliminary Analyses

We first compared the two samples on demographic, clinical, and legal factors. From a clinical standpoint, the Brooklyn and Washoe samples did not differ significantly. They had similar functioning (GAF), current symptomatology (CSI), and insight into mental illness (ITAQ) scores. When most severe diagnosis was dichotomized into schizo-spectrum disorders versus other diagnoses, the two sites also appeared similar (see Table 2).

However, from demographic and criminal standpoints, the two sites were distinct, with the exception of mean age. The Brooklyn MHC sample included significantly more men and more persons of color than the Washoe County MHC sample. The Brooklyn court, which is a felony MHC (Fisler, 2005), also included more felons and higher severity crimes than the Washoe court.

The number of days between entry date into the court and the interview did not differ significantly by site (Brooklyn M = 10.81 days, SD = 16.31; Washoe M = 13.70 days, SD = 17.40). Brooklyn participants had significantly more contact with the judge before the interview, however. On average, Brooklyn participants saw the judge 6.09 (SD = 3.54) times, whereas Washoe participants saw the judge only 2.13 (SD = 1.49) times, F(1, 195) = 113.56, p < .001, d = 1.46. As part of their standard protocol, the Brooklyn MHC has potential clients appear before the judge several times before formally enrolling in the court (i.e., entering the guilty plea). Thus, the difference in the number of times clients of MHCs saw the judge occurred before official enrollment. Nevertheless, going before the judge could influence court comprehension. Overall, because of differences between data collection sites, we either conducted analyses separately by site or included site as a factor in the analysis.

Finally, the counterbalanced order of the MacCAT-CA and the A-MHC and C-MHC did not significantly influence scores on these three measures, and thus, the counterbalance order was not entered into further analyses.

Voluntary and Informed MHC Decision

When simply asked whether entering the MHC was their choice, 66% and 71% of Brooklyn and Washoe clients, respectively, answered ‘yes.’ However, when asked whether anyone had explained to them prior to enrolling that they had a choice, 58 and 60% of Brooklyn and Washoe participants, respectively, claimed not to have been told it was voluntary. Further, when asked “After the court has said you are eligible to participate in mental health court, who makes the final decision about whether or not you participate?” the majority (73–82% depending on site) incorrectly stated someone other than themselves, most often the judge of the MHC (see Table 3). None of these rates was significantly influenced by length of time in court prior to the interview.

Table 3 MHC awareness and comprehension
Full size table

The overall mean Perceived Coercion scores (PC; 0 = low coercion; 4 = high coercion) was 1.63 (SD = 0.93). We conducted a 2 (data collection site) × 2 (yes/no: the client claimed to have been aware of the voluntary nature prior to entering) ANOVA with mean PC scores as the dependent variable. Similar to what Poythress et al. (2002) found, there was a main effect of voluntary awareness, such that those claiming to be aware prior to entering had significantly lower perceived coercion scores, M = 1.33 (SD = 0.79) in comparison to those who claimed not to have been told, M = 1.85 (SD = 0.96), F(1, 194) = 16.06, p < .0001, d = .37. The main effect for site and the interaction between site and voluntary awareness were not significant. We conducted an ANCOVA controlling for length of time in court prior to interview and found the same results.

Participants were also asked whether MHC requirements had been explained to them and if so, when. In Brooklyn and Washoe, respectively, 57 and 74% either answered that the requirements had never been explained, or explained only after they enrolled in the court (Table 3). We conducted a similar ANOVA to that described above replacing voluntary awareness with yes/no: the client claimed to have been aware of the requirement nature prior to entering. There was a main effect of requirement awareness, F(1, 194) = 6.89, p < .01, that was qualified by a significant site versus requirement awareness interaction, F(2, 194) = 4.76, p = .03. In Washoe, participants who claimed not to have been told of the requirements felt more coerced, M = 1.88 (SD = 1.05) than those who claimed to have been told, M = 1.21 (SD = 0.74), d = .83. In Brooklyn, the same pattern was not observed: for those unaware of requirements, M = 1.58 (SD = 0.79), or for those aware, M = 1.52 (SD = 0.86). We again conducted an ANCOVA controlling for time between MHC entry and interview; results were unaffected.

In an open-ended format, participants were asked what were the advantages and disadvantages of being in a MHC (participants were given two opportunities for each to answer). Almost all participants (91%) could cite advantages to being enrolled in the court, which included answers such as getting out of jail and getting into treatment. In contrast, from 46% (Washoe) to 59% (Brooklyn) could not cite a single disadvantage. Of those who could cite disadvantages, these included the time commitment, the stigma associated with being in a ‘mental’ health court, having to take medications and follow orders, and being closely monitored. A Spearman’s correlation conducted between being able to cite disadvantages (yes/no) and length of time from MHC entry to interview revealed a significant correlation, total sample r = .18, p < .01 (Brooklyn r = .12, p = ns; Washoe r = .21, p < .05). Specifically, participants, especially in Washoe, who had been in the court longer prior to the interview were better able to name disadvantages to being in the court, which may indicate that the drawbacks associated with the court are not well defined when the MHC option is presented but rather become more apparent to the client as time in the court elapses.

MHC Knowledge

When asked whether anyone charged with a crime can be in a MHC, most participants, (65% of Brooklyn and 76% of Washoe clients), correctly knew that not just anyone could enroll. However, when participants were asked whether persons can stop being in the court if he or she wanted to, 20–30% erroneously said “no” and an additional 30–34% did not know the answer (Table 3). In a follow-up question, regardless of their previous answer, participants were asked “What should a person do if she or he doesn’t want to be in the mental health court anymore?” More than half of participants knew or could at least guess at an appropriate action to remove themselves from the court, which most often reflected talking to someone like the judge, MHC coordinator, or their lawyer. However, approximately one-third of all participants (43% in Brooklyn, 21% in Washoe) either could not develop an answer to this question (i.e., “don’t know”), provided an answer reflecting an inappropriate way to stop (e.g., abscond), or claimed that a way to stop participating in the court did not exist. Again, these findings did not vary as a function of length of time in court prior to interview.

The main measure of MHC knowledge was the three series of true/false/I don’t know statements, all of which were scored as incorrect (0) or correct (1). Don’t know answers were scored as incorrect. The first series included 14 statements about what people are agreeing to when they enroll in a MHC, such as pleading guilty to the crime they had been charged with, to meet with a case manager regularly, and to not do any criminal acts. Scores on these 14 items ranged from 6 to 14; the mean score was 11.57 (SD = 1.66).

Almost all participants knew that they had to come back and see the judge (Brooklyn: 95%, Washoe: 96% correct), that they were required to take medications (Brooklyn: 98%, Washoe: 99% correct), and that they were allowing the MHC access to their health records (Both sites: 89% correct). A surprising number did not know that pleading guilty was a condition of enrollment (Brooklyn: 73%, Washoe: 55% correct). Some also did not know what was not required of them, such as having to take lie detector tests (Brooklyn: 57%, Washoe: 64% correct) and having to do physical exercise (Brooklyn: 58%, Washoe: 85% correct).

The second series included 11 statements about what can happen if people do not follow the conditions of the MHC. Scores on these 11 items ranged from 2 to 10 and the mean was 6.38 (SD = 1.54). Almost all participants were aware that they could be sent back to jail (Brooklyn: 98%, Washoe: 96% correct), though not all Brooklyn participants knew that they could be required to come back and see the judge more often (Brooklyn: 76%, Washoe: 93% correct) or be required to meet with their case managers or probation officers more often (Brooklyn: 83%, Washoe: 94% correct). In addition, some were unaware they could be kicked out of the MHC (Brooklyn: 73%, Washoe: 78% correct), or have their original charges or sentences reinstated (Brooklyn: 72%, Washoe: 81% correct). Finally, many clients erroneously believed they could be charged again for the same crime if they did not follow MHC conditions (Brooklyn: 29%, Washoe: 57% correct).

The third series included seven statements about what can happen if people do follow the conditions. Scores for these seven items ranged from 0 to 7 and the mean was 5.44 (SD = 1.30). Although there were some site differences, almost all knew that if they followed MHC conditions, they could graduate (Brooklyn: 90%, Washoe: 98% correct) and avoid going to jail or prison for their current crime (Brooklyn: 98%, Washoe: 86%). Fewer (but still the majority) knew they could get access to mental health and drug treatment (Brooklyn: 84%, Washoe: 86% correct) and have the arrest or conviction that led to MHC dropped from their record (Brooklyn: 89%, Washoe: 77% correct). Some participants wrongly believed if they followed MHC conditions, their mental health problems would be cured (Brooklyn: 39%, Washoe: 62% correct) and that they could never be arrested again (Brooklyn: 73%, Washoe: 88% correct).

An ANOVA was performed with site as the independent variable and the summary MHC Knowledge score as the dependent measure. The analysis was significant, F(1, 198) = 19.37, p < .001, indicating that Brooklyn participants were significantly less knowledgeable than Washoe participants. The mean score for Brooklyn was 22.29 (SD = 3.33), and for Washoe was 24.34 (SD = 3.43), d = .61. When ANCOVAs were performed, controlling for GAF scores, severity of primary diagnosis, and time in court prior to the interview, results of data collection site were unaffected.

Lastly, participants were asked whether they thought being in MHC would be harder, easier, or about the same as being in regular court. Overall, 18% believed it would be harder (n = 35), 46% believed it would be easier (n = 91), and 36% (n = 70) believed it would be about the same (see Table 3). We examined mean MHC Knowledge scores by this belief via a one-way ANOVA with the three levels as the independent variable and mean knowledge scores as the dependent variable. There was a significant main effect, F(2, 193) = 4.10, p = .02. LSD post hoc comparisons revealed that persons who believed MHC would be harder than regular court, M = 24.34 (SD = 3.81) and persons who believed it would be easier, M = 23.62 (SD = 3.17) were significantly more knowledgeable than persons believing MHC would be similar to regular criminal court, M = 22.47 (SD = 3.48), ds ≥ .35. Mean knowledge scores for clients endorsing ‘harder’ and ‘easier’ did not differ significantly. When time in court prior to interview was held constant, results remained the same. However, when data collection site was held constant, the significance of the main effect fell to p = .09. These findings suggest that those with less knowledge of the mental health court, including the nuances that make these courts “specialty courts,” may not understand and appreciate the differences between regular criminal court and mental health courts.

Adjudicative Competence

Legal competence was measured with the Understanding and Reasoning portions of the MacCAT-CA. The measure provides standardized norms for three classifications: 1) Minimal or No Impairment; 2) Mild Impairment; and 3) Clinically Significant Impairment. For the Understanding portion, we found the following for the Brooklyn clients: 27% demonstrated clinically significant impairments, 10% demonstrated mild impairments, and 63% demonstrated minimal to no impairment; and the following for Washoe clients: 13% demonstrated clinically significant impairments, 8% demonstrated mild impairments, and 80% demonstrated minimal to no impairment. On the Reasoning portion, 16% of Brooklyn clients of MHCs demonstrated clinically significant impairments, 25% demonstrated mild impairments, and 59% demonstrated minimal or no impairments. In Washoe, 9% demonstrated clinically significant impairments in Reasoning, 5% mild impairments, and 86% minimal or no impairments. Overall, Brooklyn clients were significantly more likely to have impairments in both Understanding and Reasoning compared to Washoe clients, χ2(2) ≥ 7.73, ps ≤ .02, Cramer’s V ≥ .20.

We also examined the relation between influence of length of time in court prior to the interview and adjudicative competence levels. Using the total sample, we found a Spearman’s r of .13, p = .06 for MacCAT-Reasoning scores (the correlation for Understanding scores was non-significant at r = .01). Upon further inspection, this trend was driven by Washoe participants; specifically, when we computed correlations separately by site, Spearman’s r between time in court and Reasoning scores was .14 (p = .16) for Washoe participants and .01 for Brooklyn participants. Thus, there was a non-significant trend indicating that Washoe participants who were in the court longer prior to their interview scored higher on the Reasoning section of the MacCAT-CA.

Predictors of MHC Voluntary Choice, Knowledge, and Comprehension

To examine the influence of demographic, clinical, and criminal factors, we performed multivariate linear regression analyses predicting MHC voluntary choice and MHC knowledge. For MHC voluntary choice, we utilized perceived coercion scores (0 = not at all coerced/voluntary to 4 = very coerced/involuntary) as the dependent measure. We entered the variables listed in Table 4 in the three steps in the order shown (i.e., Demographic and Other Factors, Clinical Factors, and Legal Competence). The model was significant, F(15, 166) = 2.38, p < .01, R 2 = .18. Clients of MHCs who went before the judge more times prior to the interview felt more coerced into enrolling than those who went less often. Persons with insight into their mental health problems were more likely to feel it was their choice to enroll in the court than those with less insight. And, similar to the results reported above, clients of MHCs who claimed not to know they had a choice felt more coerced than those who claimed to know. Adjudicative competence scores did not influence perception of voluntariness/coercion to enroll in the court.

Table 4 Logistic regression results for MHC voluntariness and MHC knowledge
Full size table

To contrast results with MHC voluntary choice, we conducted a similar regression to predict MHC Knowledge scores (Table 4). The model was significant, F(15, 166) = 6.27, p < .001, R 2 = .36. The most robust predictors of MHC knowledge were the two MacCAT-CA Understanding and Reasoning competence scores: persons who held less knowledge and appreciation of their legal situation generally held less knowledge of MHCs specifically. Other factors that predicted increased MHC Knowledge were gender (females were more knowledgeable), crime severity (persons charged with less serious crimes were more knowledgeable), and employment status (persons who were employed were more knowledgeable). None of the clinical factors, including having a schizo-spectrum diagnosis, was predictive of MHC knowledge.

Overall, the factors that predicted MHC choice and MHC knowledge were quite distinct, suggesting that the two are separate constructs. Indeed, the correlation between choice and knowledge was non-significant, Pearson’s r = .04 (non-significant for both sites when computed separately). Thus, as indicated by these regressions, the perceived choice to voluntarily enroll in a MHC is not influenced by one’s knowledge level, and knowledge is not influenced by choice.

Discussion

The goal of this study was to examine whether newly enrolled clients of MHCs made knowing, informed, and voluntary decisions to enter the court. We were interested in describing both the state of MHC comprehension at the two participating courts, as well as the demographic, clinical, and criminological factors that predicted perceived choice and knowledge. In addition, we examined legal competence. Below, we summarize and interpret findings as they pertain to voluntariness, knowledge, and competence.

On the one hand, there were indications that clients in the MHCs in this study made knowing, intelligent, and voluntary enrollment decisions. The majority (more than half): 1) claimed it was their decision to enter the MHC; 2) knew that not just any person charged with a crime could participate in a MHC; 3) were able to cite advantages to being in the court; and 4) demonstrated only minimal or no impairment on a measure of adjudicative competence (both Understanding and Reasoning). The majority also knew several “basics” of the courts, including that they had to return for judicial status review hearings, that they could be sent to jail for non-compliance, and that they were required to take their medications.

However, on the other hand, there were also indications that enrollment decisions were not knowing, intelligent, or voluntary. That is, more than half: 1) claimed not to have been told the decision to enroll in the court was voluntary prior to enrolling; 2) claimed not to have been told of the requirements of the court prior to enrolling; 3) did not know that the final decision (after eligibility decisions) to enroll in the court was theirs to make; 4) did not know they could stop being in the court if they so chose; and 5) could not cite even one disadvantage to being in the court.

We also found that 16–27% of Brooklyn and 9–13% of Washoe participants demonstrated clinically significant impairments in their understanding of legal terms and concepts, and their ability to reason pertinent to legal decision-making. An additional 5–25% demonstrated mild impairments. In addition, scores on the adjudicative competence measure and MHC knowledge were strongly related. Thus, those who did not possess a full understanding of general legal concepts or have complete reasoning capabilities also did not possess a full understanding of MHC procedures and requirements.

In theory, all clients of MHCs are presumed competent and thus allowed to plead guilty. In Godinez v. Moran (1993), the Supreme Court explicitly rejected the notion that competence to plead guilty required a higher or different standard than that articulated in the Dusky standard for competence to stand trial. Thus, our findings that a small but significant minority received scores in the clinically impaired range on a measure of adjudicative competence may be cause for concern. However, we caution against over-interpretation of the adjudicative competence findings as the MacCAT-CA is a research tool and not an actual finding of legal competence. Further, we do not know similar rates of incompetence for offenders with or without mental illness who pleaded guilty in traditional court without this threshold issue being raised.

In general, the findings in this study question whether MHCs are as truly voluntary as intended. Findings across the two courts in this study and the Broward County, FL MHC (Boothroyd et al., 2003; Poythress et al., 2002) are remarkably similar. MHCs are recent innovations (Petrila & Redlich, 2008) and remain untested, despite their prolific growth. Because MHCs are composed of individuals with serious mental illness, traditional criminal court mechanisms to gauge voluntariness (e.g., simply asking whether decisions are voluntary in plea colloquies) may not be sufficient. In this study, although many perceived that it was their choice to enroll in the court, further questioning revealed that more than half claimed never to have been told they had a choice or had been told only after enrolling. Thus, it is possible that these clients of MHCs were aware, or at least felt they had a choice, even without being explicitly told.

Despite the fact that both participating courts had written, formalized procedures to describe the courts (see descriptions in Methods section), some clients of MHCs 1) claimed not to have been informed, and 2) were inaccurate when queried about certain aspects of the court, such as having to return more often if non-compliant. Interestingly, perceived voluntariness and claims of having been informed were unrelated to MHC knowledge. Voluntariness and MHC knowledge were predicted by different factors, lending further support to the notion that these are distinct concepts in legal decision-making. For example, in Brady v. United States (1970), the Court stated “Waivers of constitutional rights not only must be voluntary, but must be knowing, intelligent acts done with sufficient awareness of the relevant circumstances and likely consequences” (p. 748). Thus, although by definition, entry into all MHCs is a voluntary decision, the decision may not be knowing and informed. Whereas decisions that are truly knowing and informed may not be voluntary, decisions that are not knowing and informed can never truly be voluntary.

Arguably, one of the more intriguing findings was that approximately half of the sample could not cite a disadvantage to being in the court, even when given two opportunities to do so. (We note that 91% of participants had the ability to cite advantages and those who could cite disadvantages cited logical and appropriate ones. Thus, we do not believe this finding is due to lack of ability.) In Washoe, this inability was negatively related to the length of time in the court, such that clients of MHCs in the court longer were more likely to cite disadvantages than those in the court a shorter time. This latter finding indicates that disadvantages, which include MHC sanctions, potential for stigma, and increased supervision, may not be well explained to potential participants prior to court enrollment, but become more apparent as participation in the court progresses.

Finally, we examined possible predictors of perceived voluntariness and MHC knowledge. The MacCAT competence scores did not significantly predict perceived voluntariness. What did predict perceptions of choice were the number of times the participant went to court prior to the interview, claims of being told the court was voluntary, and insight into mental health problems. Perhaps not surprisingly, persons with more insight into their illnesses were more likely to perceive it as their choice to enter the court. This may indicate that, in comparison to persons with more insight, those with less had to be cajoled into joining and accepting that the court was appropriate for them. Trupin et al. (2000) in studying enrollment decisions in two MHCs in Seattle, WA similarly found that defendants with more insight into their mental illness were more likely to opt in to the court than those defendants with less.

The most robust predictors of MHC Knowledge were the MacCAT Understanding and Reasoning competence scores. Participants who were females, who were charged with less severe crimes, and who were employed prior to their arrest were more knowledgeable than their counterparts. Although the employment findings are consistent with the general adjudicative competence literature, the findings concerning gender and crime severity are not. Gender does not typically influence ratings of competence (Poythress et al., 1998), and crime severity typically shows an effect in the opposite direction (such that persons with more severe crimes are less likely to have competence-related deficits; see Stafford & Wygant, 2005). However, although Cooper and Zapf (2003) had found a significant, negative relationship between likelihood of being found incompetent and crime severity, when a multivariate regression was performed, this relationship was no longer significant. It is possible that these significant effects of gender and crime severity were driven by site differences in these factors, thereby affecting MHC knowledge.

In addition, clinical factors, particularly presence of a schizo-spectrum, are usually found to be positively associated with competence impairments. In this study, generally we did not find clinical factors to predict perceived voluntariness or MHC knowledge. Although preliminary, it may be encouraging that having a schizo-spectrum diagnosis does not influence knowledge as it would neither be feasible nor desirable to exclude this group from MHCs.

Conclusions and Limitations

To our knowledge, this was the first study to examine in-depth enrollment decisions of new clients of MHCs. It was also the first to examine adjudicative competence-related deficits among these clients. Although MHCs are a new phenomenon in the realm of mental health law, they are experiencing exponential growth.

Several limitations need to be discussed. First, we did not attempt to measure what was specifically told to potential clients of MHCs or when. Rather, this was a study that examined voluntary choice and claims of being given information from newly enrolled clients of MHCs’ points-of-view. Both participating courts had standardized documents for presenting information about the court to prospective clients. What is not known, however, is the stability and severity of mental health symptoms for individual participants at the time the MHC option was explained. Second, although slightly more than one quarter of participants were interviewed on the same day as MHC enrollment, it is possible that for the other participants, the knowledge that was measured was contaminated from experiences gained in the court within the first month. However, with the exception of the ability to cite disadvantages, we did not find time in court to significantly influence MHC Knowledge or any other measures.

Third, because MHCs are notoriously idiosyncratic (e.g., Wolff & Pogorzelski, 2005), findings from these two courts may not generalize to other MHCs as the people in the court, the procedures and requirements, and use of sanctions differ. For example, some courts require participants to have a serious mental illness that influenced the commission of the crime, whereas others are vague and only require a demonstrable mental illness; some claim not to supervise clients in the community, whereas others have multiple forms of community supervision (Redlich et al., 2006). How and where court information is conveyed to potential clients is sure to vary across courts (e.g., written, discussed in person in jail). The impact of these differences on perceptions of voluntariness and comprehension is a worthy avenue for future research. On a related point, some of the answers that would be considered “correct” for these two courts may not be so for other courts (e.g., pleading guilty as a condition of enrollment). Finally, perceived voluntariness and knowledge of analogous situations (e.g., plea-taking) among persons without mental illness were not examined. However, we have a study underway to collect comparable data to better interpret this study results.

Despite these limitations, findings from this study may have important implications for mental health courts and the people they serve. MHCs, although popular, are controversial. One controversy is whether the courts are indeed voluntary as intended (Redlich, 2005; Seltzer, 2005; Stefan & Winick, 2005). Findings from this and other studies indicate that the majority of clients claim not to have known they had a choice, though maintain relatively low perceived coercion scores (Boothroyd et al., 2003; O’Keefe, 2006; Poythress et al., 2002). Although it is encouraging that there were close-to-ceiling effects for the “basics” of mental health courts, arguably individuals making important legal and treatment decisions should have more than a basic knowledge of procedures, requirements, and consequences, particularly given that there are sanctions for non-compliance. Thus, MHCs must now ask: What information do we want MHC participants to have at the time enrollment? and How can we ensure that the information is meaningfully understood, particularly the complicated nuances?Footnote 3

As a next step in this research, we are examining what in many ways is the core issue for MHCs: whether knowledge of choice and overall MHC knowledge predict MHC-related outcomes at one year, including arrests and violations, and success in the court. If there is a negative relation between knowledge at onset and future success, such that clients of MHCs who lack comprehension do not succeed (e.g., engage in treatment, take medications, are re-arrested, etc.), MHCs will be less likely to realize their goals of reducing the cycle of arrest and incarceration for people with serious mental health problems. The extant research on MHCs suggests that this is not an intervention that works equally well for all, and research is now aimed at determining the factors that are predictive of success. Sufficient MHC comprehension at enrollment may be such a factor.