As the census in forensic psychiatric hospitals grows, psychiatrists are under increasing pressure to develop strategies to address the problem of coexisting criminal behavior, addiction, and mental illness. In particular, effective treatment is hindered by the paucity of research on clinical application of treatment models for addictive disorders in this population. This article reviews the literature on the complex relationship among violent behavior, psychiatric illness, and substance use, and considers the unique treatment needs and opportunities for intervention among forensic psychiatric inpatients.
Epidemiologic studies have demonstrated a link among mental illness, violence, and substance use. Using Epidemiologic Catchment Area data on 10,000 subjects, Swanson et al.1 found that respondents with affective and schizophrenic disorders had similar rates of assaultive behavior in the past year (10.7%-12.7%), whereas those with no psychiatric diagnosis had a rate of 2.1%. Conversely, the prevalence of major mental illness was 3 to 4 times higher among violent respondents than among nonviolent respondents. However, respondents with substance abuse diagnoses had rates of assaultive behavior that ranged from 19.3% to 34.7%, and accounted for almost half of the violent behavior.
There was a linear relationship between the number of diagnoses and the rate of violence. Comorbid addiction diagnoses accounted for the higher rates of violence in affective disorders, and for much of the increased violence rates in schizophrenia. Major mental disorder, addictive disorders, and addiction in combination with major mental disorder were found to be significant predictors of violent behavior, with substance abuse or dependence being the single most important predictor.
Three longitudinal prospective studies of unselected Scandinavian birth cohorts have yielded further information on the relationship among psychiatric illness, substance use, and violent behavior. Ortmann et al.2 observed a male Danish birth cohort through their early 20s and found that 35% of all men without a diagnosed major mental illness had some history of police registration, compared with 44% of those with a major mental illness and 83% of those with a documented substance-related problem.
Hodgins3 examined an unselected Swedish birth cohort observed to age 30. Men with major mental disorders were 2V¿ times more likely than men with no disorder to commit a crime, and 4 times more likely to commit a violent offense. Women with major mental disorders were 5 times more likely than women with no disorder to commit a crime, and 27 times more likely to commit a violent offense. Almost half of the offenders with a major mental disorder had comorbid substance use problems.
Tiihonen et al.4 analyzed data from a male Finnish birth cohort to determine the quantitative risk of criminal behavior associated with specific mental disorders. Those with a major mental disorder had odds ratios from 3 (schizophrenia) to 30 (organic mental disorders) of committing a crime compared with those without a major mental disorder. The risk of committing a crime was 4 times higher for alcoholic schizophrenics than for nonalcoholic schizophrenics. However, when examining these Scandinavian birth cohorts, it is important to keep in mind that data may not be generalizable to countries such as the United States that have high crime rates, especially high crime rates by substance abusers.
Other bodies of research have examined the interactions among psychiatric disorder, substance use disorders, and violence from different angles. Many studies have focused on violent behavior in psychiatric patients,5"7 resulting in widely varied estimates of violence in this population; this variability largely reflects differences in diagnostic mixes and acuity levels. One review8 suggests that schizophrenic individuals are at higher risk (and individuals with affective disorder are at lower risk) for acting violently than are those with other diagnoses, but substance abuse or dependence has been an overlooked correlate of violence in many studies. Attempting to clarify the role of psychoactive substance use in assaultive behavior by schizophrenics, Lindqvist and Allebeck9 identified a cohort of people hospitalized for schizophrenia and examined records of crimes committed by those individuals in subsequent years. They found psychoactive substance use to be a problem in more than half of the violent schizophrenics.
Yet another line of research has assessed psychiatric disorders in violent criminals, consistently finding the prevalence of mental illness to be higher among offenders than in the general population.10,11 Yarvis12 examined patterns of psychoactive substance use among murderers and found that more than half of the offenders were using at the time of the crime. Although psychoactive substance use was deemed to play an insignificant role in murderers who were psychotic, it was a key factor in offenders with nonpsychotic Axis I diagnoses and those with severe Axis ? pathology.
Taken together, these studies suggest that the mentally ill are somewhat more likely than the general population to commit crimes, but the risk of violence posed by mental illness is much less than that posed by addictive disorders. Also, the elevated risk conveyed by severe mental illness is substantially mediated by comorbidity with psychoactive substance use disorders in this population.
Forensic psychiatric inpatients comprise a population distinct from the general psychiatric population, prisoners, and the population at large. The forensic group has unique diagnostic profiles, treatment needs, and opportunities for intervention. Each of these factors will be examined to determine the best treatment strategy for forensic patients with concurrent psychoactive substance use or dependence.
With few exceptions, forensic patients fall into one of two categories- incompetent to stand trial (1ST) or not guilty by reason of insanity (NGRI). 1ST defendants most commonly carry a diagnosis of a psychotic disorder, and frequently manifest bizarre behavior and impaired orientation13- as might be expected of those recently adjudicated as currently unable to assist in their defense. Factors associated with a lack of criminal responsibility (NGRI) are similar, including bizarre behavior at the time of the offense and a current diagnosis of a psychotic disorder.13 Thus, although there may be demographic and behavioral differences between the groups- with 1ST patients tending to be younger and to have a more pronounced history of criminal behavior- diagnoses of schizophrenia or severe affective disorders, especially bipolar illness,13 are common among both populations.
Rather, differences in the clinical status of 1ST versus NGRI patients often manifest more in degree of current stability. The delay between the occurrence of the offense and the judicial system's decision to remand the patient to a facility for treatment after being found NGRI is typically lengthy, often with an adjudication of 1ST and treatment for restoration of competence intervening. Thus, NGRI patients tend to have multiple opportunities before trial, either in jail or while being held as 1ST, for an acute psychiatric disturbance to be identified and treatment initiated. Only after the defendant is actually tried and acquitted as not criminally responsible does the (usually lengthy) hospitalization as an NGRI patient begin.
Psychoactive substance abuse or dependence is a frequent comorbid condition in schizophrenia and severe mood disorders among both forensic and nonforensic patients14 and can play a significant role in terms of clinical course and treatment, although not necessarily in legal disposition. But it is important to address substance-related problems in the context of forensic treatment, in part because it influences the ultimate success of NGRI discharges. As noted by Fitch et al.,15 most revocations of conditional release from the inpatient setting are the result of drug or alcohol use. Our experience is similar, with more than half of failed conditional discharges at our institution being directly related to substance use (ie, using illicit substances in violation of the terms of discharge or exhibiting decompensation secondary to relapse of addiction).
Forensic hospitalization may provide a singular opportunity for intervention in the course of an addictive disorder. Because the NGRI population has a long average length of stay (generally measured in years), it makes clinical sense to incorporate treatment of the addictive disorder as an integral part of therapeutic programming. The fact that relatively stähle patients may be retained in a highly structured setting for an extended period of time permits the use of intensive, long-term programs that would not be feasible in other contexts. Moreover, the highly selected nature of this treatment population suggests that such interventions might have a disproportionate impact in reducing violence among a population with other significant, nonmodifiable risk factors. NGRI patients' awareness that the discharging authority considers the hospital course in making discharge decisions may motivate them to actively participate in addiction programming. The discharging authority has further potential leverage because it can require such treatment as part of any disposition.
Mr. A is a 41-year-old, never-married African-American man who was adjudicated as NGRI in 1981 following trial for the murder of his stepfather. His psychiatric history began 2 years earlier, when he developed sustained low mood (for at least 6 months prior to the crime), accompanied by a full range of depressive symptoms. At the time of the murder, he was unemployed and living in a basement apartment of a building where his mother and stepfather resided.
He was also drinking up to two six-packs of beer plus a pint of rum per day, as well as taking diazepam. In retrospect, he also felt that about that time his mind was "playing tricks" on him and he was "getting paranoid," and had begun carrying a knife for self-protection. His family described him as then being depressed, suspicious, and withdrawn, and in fact had scheduled a psychiatric evaluation for him.
On the day of the crime, Mr. A felt "funny" and dysphoric. He drank heavily, smoked marijuana, and that evening joined his stepfather and watched television, noting that the stepfather "was looking at me like he was going to do something to me." He returned to his apartment, drank more, and rejoined his stepfather and mother. When Mr. A's mother asked him to turn off the television, he felt that "something inside me said to stab" the stepfather, and did so. After the stabbing, he ran downstairs and told an uncle what he had done. The police were called, and, on questioning, Mr. A denied any animosity toward the victim.
Mr. A was diagnosed as suffering from major depression, with mood-incongruent psychotic features (later changed to schizoaffective disorder, depressive type) as well as alcohol dependence, and was hospitalized following acquittal. Although initially uncooperative, once he agreed to treatment he rapidly improved and conditional release was granted on the second petition.
However, his conditional release was revoked after approximately 6 months due to resumption of problematic alcohol use. As during his first hospitalization, Mr. A reported persecutory delusions and initially refused treatment, but rapidly improved after agreeing to restart medication. However, despite stating that his ongoing desire for alcohol is "in my genes," he has declined to participate in any form of addiction treatment. At the time of this writing, no petition for release had been placed before the court.
Mr. A's case demonstrates several features common in this population: (Da psychotic illness complicated by coexistent addiction (to alcohol in this case); (2) evidence of both intoxication and psychotic features active at the time of the crime, yet with insufficient attention initially paid to the addiction by treating professionals; (3) relapse of the psychiatric illness, perhaps first heralded by resumption of substance misuse; and (4) likely both failure of outpatient treatment and prolongation of hospitalization due to unsuccessful treatment of his addictive disorder. Without adequate intervention, the prospects for successful long-term treatment are remote.
INSANITY VERSUS INCOMPETENCE TO STAND TRIAL
As compared with insanity acquittées, IST populations tend to be younger, to be, less stable psychiatrically, and to have a substantially shorter average length of stay. Rather than focusing on extensive rehabilitation programs, with a goal of reentry into society, the treatment goal for 1ST patients is primarily restoration of competence to stand trial. If trial results in a verdict of NGRI, the patient may be remanded to the facility for more extensive treatment, but otherwise the patient will generally be imprisoned (if convicted) or released (if acquitted); in either case, there is little opportunity for more extensive therapy, at least in an inpatient setting.
This difference in orientation brings with it a vastly different set of motivations and treatment considerations. During the initial phases of their hospital stay, patients admitted as 1ST are more likely to demonstrate marked symptoms of their underlying psychiatric illnesses; only after these have begun to abate is treatment for comorbid addictive disorders indicated. With no clear benefit from this intervention likely at trial, and because restoration to competence may soon be followed by a prison term, patients adjudicated 1ST are less likely to be motivated to participate in treatment. In addition, any course of treatment may be terminated by a hearing that determines that restoration has been accomplished, leading to return to jail pending trial.
Given the significant differences between the groups, it is unfortunate that there is little information to guide clinicians as to what types of addiction treatment might be most efficacious. It is likely that most institutional programs rely on abstinence-based, 12-step programs, either alone or as a major component of treatment, regardless of coexisting psycho pathology- as is true in nonforensic settings as well. Although there is good evidence from community-based treatment centers that such programs may be effective even when full participation is lacking, it is also clear that for optimal results, continuation in formal aftercare programming and other post- treatment factors is required16- potentially a difficulty for forensic patients, particularly those adjudicated 1ST. Moreover, although there is evidence (again primarily from community-based rather than forensic treatment settings) that such abstinence-based programming can be effective when integrated into ongoing inpatient treatment of individuals with severe mental illness, for optimal effect such programming requires adaptation to the special needs of the patient, including intensive case management following discharge.17·18 Despite their demonstrable clinical appropriateness, such services are often not easily available to those adjudicated 1ST Conversely, among NGRI acquittées, abstinence after discharge may prove more attainable (if this goal is sufficiently promoted during the inpatient stay) by virtue of their lengthy hospitalizations and the consequent enforced abstinence.
There is also some evidence to suggest that matching interventions to specific patient characteristics may be of benefit. McLellan et al.19 reported on a program matching substance abuse services to targeted individual needs in the areas of employment, family, and psychiatry. They found that matched patients stayed in treatment longer, were more likely to complete treatment, and had better 6-month outcomes than standard-care patients in the same program. Similarly, Litt et al.20 identified two alcoholic subtypes- type A (characterized by later onset, fewer childhood risk factors, less severe dependence, fewer alcohol-related problems, and less psychopathological dysfunction) and type B (marked by sociopathy and perhaps more resembling alcoholic 1ST patients). They found the latter had better outcomes with coping skills training and responded less well to interactional therapy.
Thus, there is some evidence (although derived from nonforensic samples) that clinically relevant subtyping of patient characteristics and needs may improve compliance and outcome. For NGRI patients, there is more opportunity for lengthy, intensive interventions that can target specific areas of need in domains such as employment skills, family relations, and personality styles following a substantial period of psychiatric stabilization. Thus, a more individualized, intensive, long-term approach is both practical and potentially more likely to result in remission than a standardized, time-limited program, particularly if during the course of treatment efforts are made to address relevant post-treatment factors such as continuing care and participation in 12-step programming, if used.
By contrast, 1ST patients, whose stay at a facility is typically much briefer and whose psychiatric illness as a result tends to be less well controlled, are unlikely to have the opportunity (and perhaps the motivation) to participate in such intensive programs. This population may differentially respond to more structured interventions that concentrate more on learning specific coping skills as well as dealing with the denial of their addictive illness and the need to comply with treatment interventions. Rather than being geared toward a definitive outcome, treatment efforts might better be conceptualized as being focused primarily on engaging the patient in the idea of treatment and providing for its initial stages.
One of the biggest obstacles to treating addicted forensic patients is the lack of data to guide the treatment decisions of the forensic clinician. Those treating addicted forensic patients need to consider the following:
* Treat the coexisting psychiatric illness.
* Treat the coexisting addictive illness, keeping in mind that to be effective, treatment methods may need to be modified to take into account impairment secondary to psychiatric illness.
* There is some evidence that the 1ST population might do better with addiction treatment focused on coping skills training and on engaging them in the initial stages of treatment.
* Violence potential needs to be closely monitored and aggressively treated; of particular importance is the relationship of violence to the associated psychiatric and addictive disorders.
* Judicial and parole or probation systems need to be enlisted as therapeutic allies; emphasis should be on mandating compliance with psychiatric treatment, which may lead to a decreased risk of relapse of alcohol and drug addiction.
Clearly, it is also essential that forensic research better delineate treatments specifically geared to the NGRI and the 1ST populations. The characteristics of forensic psychiatric patients make them groups distinct from both prisoners and other psychiatric patients, with unique treatment needs and opportunities. Until more specific information is available, forensic psychiatrists will be forced to modify existing treatment approaches and extrapolate from studies of populations that may differ in fundamental ways from the group they are charged with treating.
1. Swanson JW, Hölzer CE, Ganju VK, Joño RT. Violence and psychiatric disorder in the community: evidence from the Epidemiologic Catchment Area surveys. Hospital and Community Psychiatry. 1990;41:761-770.
2. Ortrnann J. Psykisk Ofvigelse og Eriminel Adfaerd en Under Sogelse at 11,533 maend fodt i 1953 i det metropolitane omrade K?benhavn. Copenhagen, Denmark. Rapport til Justitsministeriet, Direktoratet for Kriminel Forsorgen. 1981;Fornsningsrapport 17.
3. Hodgins S. Mental disorder, intellectual deficiency, and crime. Arch Ceo Psychiatry. 1992;49:476-483.
4. Tiilionen J, isohamil M, Rassnen P Koiranea M, Muring .1. Specific major mental disorders and criminality: a 26year prospective study of the 1966 Northern Finland birth cohort. Am ,J Psychiatry. 1997;154:840-845.
5. Tardif K, Sweilam A. Assaultive behavior among chronic inpatients. Am J Psychiatry. 1982;139:212-215.
6. Craig TJ. An epidemiological study of problems associated with violence among psychiatric inpatienta. Am J Psychiatry. 19S2;139:1262-1266.
7. Thrcliff K, Sweilam A. Assault, suicide, and mental illness. Arch (len Psychiatry. 1980;37:164-169.
8. Krakowski M, Volavka J, Brizer D. Psychopathology and violence: a review of the literature. Camp Psychiatry, 1986;27:131- 146.
9. Lindqvist F, Allebeck P. Schizophrenia and assaultive behavior: the role of alcohol and drug use. Acta Psychiatr Scand. 1969;82:191.195.
10. C?td 0, Hodgins S. The prevalence of major mental disorders among homicide offenders. mt J Law Psychiatry. 1992; 15 : 89-99
11. Eronen M, Hakola P, Tiihonen J. Mental disorders and homicidal behavior in Finland. Arch (len Psychiatry. 1996;53:497-501.
12. Yarvis RM. Patterns of substance abuse and intoxication among murderers. Bull Am Acad Psychiatry Law. 1994; 22:133-144.
13. Daniel AE, Beck NC, Herath A, Schmitz M1 Menninger K. Factors correlated with psychiatric recommendations of incompetency and insanity. J Psychiatry Law. 1984; 12:527-544.
14. Regier DA, Farmer ME, Rae DS, et al. Comorbidity of mental disorders with alcohol and other drug abuse. JAMA. 1990;264:2511-2518.
15. Fitch WL, Barton CD, Wise BF. Correlates of success on conditional release Presented at the 28th Annual Meeting of the American Academy of Psychiatry and the Law; Denver, CO; 1997.
16. Miller NS, Hoffman NG, Ninonuevo F, Astrachan BM. Lifetime diagnosis of major depression as a multivariate predictor of treatment outcome for inpatients with substance use disorders from abstinence-based programs. Ann Clin Psychiatry. 1997;9:127-137.
17. Miller NS. Psychiatrie comorbidity: occurrence and treatment. Alcohol Health and Research World. 1994; 18: 261-264.
18. Drake RE, McHugo GH, Noardsy DL. Treatment oí alcoholism among schizophrenic outpatients: four year outcomes. Am J Psychiatry. 1993;150:328-329.
19. McLellan AT, Grissom GR, Zanis D, Randall M, Brill P, O'Brien CP. Problem-service "matching" in addiction treatment. Arch Gen Psychiatry. 1997;54:730-735.
20. Litt MD, Babor TF, DelBoca FK, Kadden RM, Cooney NL. Types of alcoholics: II. Application of an empirically derived typology to treatment matching. Arch Gen Psychiatry. 1992;49:609-614.