Best Practices - Concurrent Mental Health and Substance Use Disorders

Best Practice for the Treatment/Support of People with Concurrent Disorders

This section outlines the best practice guidelines for treatment/support within four sub-groupings of people with concurrent disorders:

• co-occurring substance abuse and mood and anxiety disorders;
• co-occurring substance abuse and severe and persistent mental illness;
• co-occurring substance abuse and personality disorders;
• co-occurring substance abuse and eating disorders.

Before dealing with each sub-group separately, some of the common elements of an effective approach to treatment and support are highlighted:

• both substance abuse and mental health problems can be chronic, recurring health problems which usually require some immediate interventions as well as ongoing monitoring and support. Mental health symptoms may have an independent trajectory that may nevertheless be exacerbated by substance use. This is sometimes akin to 'kindling' effect whereby the use of substances exacerbate mental health problems, which in turn exacerbate housing difficulties, problems with the law, etc. Alternatively, the mental health symptoms may be a clear consequence of chronic and heavy use of psychoactive substances (e.g., chronic depression). One purpose of ongoing monitoring and support is to help sort out the interaction between the substance use/abuse and the mental health problems and symptoms;
• consistent with our earlier definitions of program- and system-level integration some specific interventions for some sub-groups may best be delivered sequentially, while others are best delivered at the same time. Integration is about communication, consistency, and coordination of all the various clinicians and support workers and not whether one set of problems (mental health or substance abuse) is addressed before the other;
• treatment and ongoing support for all sub-groups of people with concurrent disorders should also include attention to the person's basic needs for housing, access to food, social support and other aspects of psychosocial circumstance and social functioning;^64,225
• the research evidence suggests there is considerable value in tailoring the intervention for people with concurrent disorders to the motivational level or 'stage' that the consumer is at, at that particular point in time.^62,218 Thus, for all sub-groups there may be value to assessing motivational stage and utilizing motivational interviewing techniques,²²⁶ recognizing that with some disorders clinicians may expect expressed or felt motivation to be unstable;
• there is little evidence in support of residential treatment over intensive outpatient care;³⁵
• self-help groups such as AA and other 12 Step programs^226,227,228 and consumer/survivor initiatives in the mental health field⁷⁴ play a critical role in community mental health and addiction systems.

Co-Occurring Substance Abuse and Mood and Anxiety Disorders

a) Prevalence and Etiological Issues

Several reviews in recent years have demonstrated marked comorbidity between substance use and mood and anxiety disorders.^54,229-231 Elevated associations between unipolar depression have also been consistently observed ²³¹ These reviews summarize a substantial body of evidence pointing to the high prevalence of mood and anxiety disorders among people in treatment for substance use disorders as well as elevated rates of substance abuse among individuals in mental health settings.^229,232-235

Significant rates of comorbidity between substance use disorders and mood and anxiety problems have also been found in large multi-site community studies ⁷⁵ This suggests that the positive association between these problems in clinical groups is not an artifact of self-selection into treatment. For example, the ECA study⁷⁵ found anxiety disorders to represent the most prevalent non-substance use diagnoses (19%) among people with alcohol use disorders. Those with any anxiety disorder had a 50% increase in the odds of being diagnosed with an alcohol use disorder. A similarly elevated risk was found in The National Comorbidity Survey.⁷⁶ These patterns can also be found in the community studies of mood disorder and substance abuse where lifetime prevalence of major depression among substance users was much higher (24.3%) than the rate in the general population (5.8%).⁷⁵ Furthermore, rates of substance abuse disorders among those with depression (27.2%) and bipolar disorder (60.7%) were also quite high.

The strength of the association between specific anxiety-related diagnoses and substance use disorders varies considerably. Phobic anxiety disorders and, in particular, panic disorder with agoraphobia and social phobia, appear to be most highly associated with alcohol use disorders. In addition, the overlap between post-traumatic stress syndrome (PTSD)* and substance use disorders is extremely high with estimates of co-occurrence ranging from 12%-59%.²³⁷ The prevalence of post- traumatic stress syndrome amongst people with substance use disorders is higher for women than men and estimated to range from 30-59%.²³⁶

The growing awareness of substance-induced mood and anxiety disorders⁴⁶ suggests that for many people the effective treatment of their substance abuse would alleviate the symptoms of the mood and anxiety disorders.

* Post-traumatic stress disorder is an anxiety disorder which involves a cluster of symptoms characterized by a strong tendency to avoid emotion. Post-traumatic stress syndrome may occur as a consequence of experiencing a severe and stressful life event. Simple post-traumatic stress syndrome may result following a single traumatic event whereas complex post-traumatic stress syndrome arises from repeated incidents of trauma and is associated with a broader range of symptoms.

b) Implications for Screening and Assessment

Screening for an alcohol or drug problem among people with anxiety disorders can be accomplished with one or more of the screening tools identified in Section 3.1.2 (e.g., the Alcohol Use Disorder Identification Test;¹⁸⁸ Drug Abuse Screening Test.)¹⁸⁵ Levels of severity can be rapidly determined using the Alcohol Dependence Scale²³⁷ and a more detailed picture of alcohol or drug use can be obtained with the Timeline Follow-back method.^208,216 The Beck Anxiety Inventory^,238 or the Brief Symptom Checklist¹⁹¹are useful screening instruments for mood and anxiety disorders. Specialized instruments are available to assess trauma symptoms and dissociation. One such instrument is the Trauma Symptom Checklist-33 that measures long term sequelae of sexual abuse.²³⁹ Psychiatric diagnosis of the anxiety or alcohol use disorder should be carried out by a psychiatrist, registered psychologist (in some jurisdictions also registered clinical social workers) using a structured and validated interview such as the SCID for the DSM-IV.²⁰⁵

A greater challenge is encountered in assessing the relationship between an alcohol and a mood and anxiety disorder. Such an assessment should include several sources of data including:

• family history of both disorders;
• natural history of both disorders (i.e., onset, course, fluctuations, remissions);
• the effect of improvements of one disorder on the course of the other;
• childhood variables (e.g., trauma, neglect, abuse); and
• patterns and severity of alcohol consumption (e.g., episodic vs. daily).

Assessment does not occur only at intake; re-assessment over the course of care and support is critical to sorting out whether the mood and anxiety problems are substance-induced. In the case of mood and substance abuse, Brown et al.²³⁵ and Schuckit & Monteiro⁹⁶ have suggested waiting several weeks prior to making a definitive diagnosis.

c) Clinical Issues and Treatment Implications

One of the more salient issues facing treatment of individuals with anxiety or mood disorders is the difficulty in accurately diagnosing the nature of the relationship between these disorders and concurrent substance abuse. The observation that psychoactive substances can produce signs and symptoms strongly resembling anxiety or mood disorders creates the dilemma of not knowing whether such disorders should be treated directly or if they would resolve following reduction in substance use. Also, withdrawal from many substances may also produce signs and symptoms that resemble mood or anxiety disorders which can hamper treatment. This diagnostic confusion can lead to inappropriate or delayed treatment.

Additional issues relevant to concurrent dependence on substances and mood/anxiety disorders include the increased danger of suicide, especially in the case of mood disorders. There may also be increased ambivalence evident in the motivation to address the substance problems since in many cases the abused substance may be perceived as temporarily improving the distressing mental health symptoms. It is thus very important to regularly monitor and strengthen motivational status. Furthermore, concurrent mood/anxiety disorders and substance use appear to be associated with poorer response to both psychological and pharmacological treatments;^240-241 poorer prognosis (i.e., higher relapse rates, residual dysfunction), and higher rates of medication use;²⁴² and greater probability of requiring additional treatment.²³⁵

The co-occurrence of post-traumatic stress syndrome and substance abuse is associated with poorer treatment outcomes.^243-246 People with substance use disorders and post-traumatic stress syndrome are:

• more likely to abuse harsher classes of psychoactive substances, such as heroin or cocaine;
• have high rates of substance abuse; and
• are more vulnerable to re-traumatization.²⁴⁷

Studies comparing individuals with post-traumatic stress syndrome and substance abuse disorders compared to those with substance use disorders alone have found that the former have a more impaired clinical profile. Problems include higher rates of unemployment, greater social and family dysfunction, poorer medical and psychological status, including more co-morbid Axis I disorders.^243,246

d) Review of the Evidence

This section summarizes the evidence relevant to two questions:

• Is there evidence for or against a particular sequencing of interventions?
• Is there evidence that points to one or more specific treatment approaches or interventions as being particularly effective?

i) Sequencing:

With respect to the first question, there are several compelling reasons why a sequenced approach, specifically dealing with the substance related problems first, may be more appropriate than a concurrent approach for the majority of consumers with co-occurring mood and anxiety and substance use disorders. This sequenced approach would still be undertaken within an integrated program and/or system whereby the various clinicians and support workers would have agreed to work together on this general treatment plan.

Some of the reasons for the sequenced approach include:

• the mood and anxiety syndromes that are observed in people who are alcohol dependent are often organic consequences of chronic alcohol use and can improve significantly with a reduction in use; ²⁴⁸
• psychoactive substances negatively impact on many areas of the individuals' life such as work, health, interpersonal relationships, finances and cognitive functioning. Reducing substance use will lessen the strain on these areas. For example, most of the incidents of suicide in a study by Murphy²⁴⁹ were committed under acute intoxication.^159,250 As long as psychoactive substances are still being used, the person may experience many of the consequences that are generally associated with mood/anxiety disorders.
  
  The suggestion to start with the substance abuse for most people with concurrent mood and anxiety and substance use disorders is strongly supported by a recent study by Verheul et al.¹⁴⁵ Their results show that people who were recovering from an alcohol use disorder were 16.7 times more likely to recover from their mood/anxiety disorder than people who did not recover from their alcohol use disorder. Those recovering from an opioid use disorder, were 4.3 times more likely to recover from their mood/anxiety disorder than those who did not. This increased probability was not observed among those recovering from cocaine use disorder in comparison to those who did not.
  
  While a general rule of thumb may be to begin with the substance use problem this is not likely to be successful in all cases, and an ongoing assessment will be required to adjust the treatment plan if the mood and anxiety symptoms are not alleviated through a reduction in substance use. Symptoms of anxiety and depression may not only interfere with optimum outcomes from substance abuse treatment, but are frequently reported as triggers for relapse.^251-256 If the client succeeds at significantly reducing their alcohol and/or drug consumption, and continues to experience anxiety symptoms, then there is a high likelihood that they also suffer from an independent mood or anxiety disorder and treatment should proceed to deal specifically with that disorder. Thus, comprehensive and repeated assessment is critical when a sequenced approach is followed. For other people, mood and anxiety disorders can precede and be functionally related to the substance use problem.⁵⁴ Such situations may require explicit treatment of the mood and anxiety disorder from the outset, keeping in mind that the symptoms which may be self-medicated by the substances are still exacerbated by substance misuse.
  
  There is one major exception to the general rule of thumb to start first with the substance use disorder for concurrent substance use and anxiety related disorders. This exception concerns co-occurring post-traumatic stress syndrome, whereby prevailing clinical wisdom would recommend concurrent treatment interventions within a fully integrated model.¹⁵⁹ For example, Najavits et al.²⁵⁷ argue that specific treatments known to be effective for post-traumatic stress syndrome may not be helpful on their own for individuals with concurrent substance abuse and post-traumatic stress syndrome . An example would be exposure procedures that can increase emotional intensity, which in turn may trigger a relapse to substance use. Similarly, 12- step programs often include a focus on the present rather than the past (e.g., public disclosure and surrendering to a higher power) and can be experienced as re-traumatizing to the client with post-traumatic stress syndrome.²⁵⁷ A concurrent integrated approach is also favored by clients.²⁵⁸ Although there is limited empirical evidence on concurrent versus sequenced approaches to substance use and mood and anxiety disorders, the evidence and the reasoning is strong enough to suggest such an approach be followed for post-traumatic stress syndrome.

ii) Interventions:

Whether the interventions are delivered in sequence or concurrently, what are the treatment approaches and interventions for concurrent mood and anxiety disorders and substance abuse that are the most highly supported by research? Recent reviews, summarizing a large number of well-controlled treatment studies have found overwhelming evidence for the efficacy of cognitive-behavioural therapy (CBT) for alcohol use disorders.^12,259 This includes several key components such as:

• ensuring client choice;
• treating the context of substance misuse;
• tailoring to the person's stage and motivation for change;
• practical problem-solving;
• emphasis on action;
• reliance on social supports;
• resolving ambivalence about change;
• identifying and managing cues to misuse.

Concurrent anxiety disorders:

CBT is by far the most effective treatment for anxiety disorders,^260-262 and is consistent with the CBT approaches found to be effective for alcohol use disorders. An individual whose substance use is a means of controlling their anxiety symptoms will require a strategic and effective intervention for these symptoms during the window of opportunity afforded by a reduction in drinking or drug use. Otherwise the risk of relapse and dropout may be high.²⁵³ Thus, the clinician should be able to effectively administer CBT treatments for the phobic disorders which are characterized primarily by behavioral avoidance (i.e., social phobia, agoraphobia, specific phobia), and the anxiety states which are characterized primarily by somatic activation and cognitive symptoms (e.g., panic disorder, generalized anxiety disorder, acute stress disorder). For each of these disorders there are well-established and empirically supported treatment protocols which guide the clinician in case conceptualization and treatment, including pharmacological approaches.* ²⁶⁰ Psychological treatment typically stresses:

• problem definition;
• functional analysis of the symptoms;
• identification of cognitive distortions and schemas;
• exposure to the phobic stimulus;
• arousal management; and
• other strategies.

CBT for alcohol dependence may need to take into consideration the impact of the concurrent anxiety by negotiating treatment goals; identifying the types of high-risk situations; the role of cognitive distortions; the kinds of environmental and social support and resources available; the rate of change, etc.

Concurrent mood disorders:

Similarly, with respect to mood disorders an analysis of the depressive symptoms will identify the specific cognitions, behavioral pattern and coping skills that may be maintaining the depressed mood and associated symptoms. Well-described and detailed protocols exist outlining the specific techniques that can modify depressed mood and behavior, including pharmacological treatment.^263,264 For concurrent depression and alcohol use disorders the evidence also points to CBT as an effective approach. For example, Brown et al.^265,269 treated depression and alcoholism with eight sessions of either CBT or relaxation therapy and found superior effects for the CBT treatment on mood and anxiety symptoms and percentage days abstinent. Results were even more pronounced at the 6-month follow-up.

Despite addressing the alcohol and mood/anxiety symptoms, some individuals may require ongoing relapse prevention, case management and booster sessions. These would include the clients who have difficulty coping with chronic stress or negative life events; occasionally succumb to drinking cues; display less than effective interpersonal behavior. For this population, treatment may often be an ongoing process and require the use of a wide variety of modalities and services (e.g., family therapy, vocational counseling, stress management, lifestyle re-education), in addition to direct treatment of their alcohol abuse and mood/anxiety symptoms. Individuals with concurrent personality disorders often fall into this category.

For individuals who demonstrate serious impairment in several areas, such as work or school, family relations, judgment, thinking or mood, a close and ongoing monitoring of functioning is required as the recovery from substance use disorder may not eliminate the serious impairment in functioning. These individuals remain particularly vulnerable to problematic life events due to their poor coping skills. Suggesting and enabling them to come for additional support when they experience significant stress will probably be a useful preventive strategy. With those individuals who have significant impairments in functioning, considerable vigilance should be exercised for those whose substance abuse is not improving. For them, the various risks associated with mood/anxiety disorders remain. For example, as noted earlier, the risk of suicide under acute intoxication is high.^249,250 There is also an increased risk of domestic violence,¹¹⁵ child neglect,¹¹³ child abuse, etc.¹¹⁴

Concurrent post-traumatic stress syndrome:

The limited scientific literature on the treatment of co-occurring post-traumatic stress syndrome and substance abuse does not allow for an empirically validated recommendation on the best standard of care beyond the recommendation noted earlier for a concurrent integrated model. Until recently there were no programs designed to treat post-traumatic stress syndrome and substance abuse. Recently, there have been developments in the area of integrated approaches. These researchers agree on the need for an initial phase that helps the person stabilize and improve functioning.²⁶⁶ It is also recommended that treatment help the individual learn about both disorders, including their inter-relationship and symptoms.

Najavits et al.²⁴⁶ and Najavits²⁶⁷ were the first to develop and evaluate a cognitive-behavioral approach within a concurrent, integrated model. Participation in their program, known as "Seeking Safety", was associated with high retention rates, and reduced substance use as well as post-traumatic stress syndrome symptoms.²³⁶ This treatment program is probably the most widely studied for this population and is currently being evaluated in eight different subgroups for concurrent post-traumatic stress syndrome and substance use disorder. The "Seeking Safety" treatment is designed for clients in the first stage of recovery in which the goal is to reduce substance use and post-traumatic stress syndrome symptomatology. The treatment seeks to:

• increase clients' knowledge of both disorders;
• enhance life structure and increase coping skills in the management of painful affect; and
• enhance self-care and interpersonal relationships.

Evans and Sullivan¹⁵⁹ offer another dynamic approach to addressing post-traumatic stress syndrome patterns and substance abuse.

Drug Therapies:

There is a small literature on pharmacologic treatment tailored specifically to people with concurrent mood/anxiety and substance use disorders. While some alternatives are beginning to show promise there is insufficient evidence at present for a best practice recommendation. Fluoxetine has been shown to possess relaxing effects in people with substance use disorders that were also diagnosed with social phobia and panic.^268,269 Tollefson, Montague-Clouse & Tollefson²⁷⁰ found that buspirone reduced alcohol use among people with concurrent anxiety disorders. Kranzler et al.²⁰⁴ found similar results. The use of benzodiazepines is generally cautioned against due to the risk of cross-addiction²⁰ unless the individual has been abstinent for a stable period. While traditional tricyclic anti-depressants effectively treat the depression, they do not impact on alcohol consumption.^271,272 Fluoxetine (a selective serotonin re-uptake inhibitor) reduced mood symptoms and alcohol consumption in people with alcohol dependence who were clinically depressed.²⁷³ Kranzler et al.²⁷⁴ found similar results for fluoxetine and naltrexone. There is some evidence for the use of naltrexone in the management of alcohol abuse^275-278 for people with concurrent mood/anxiety disorders.

In the past decade there has also been some evidence that pharmacotherapeutic interventions can be effective in reducing symptoms associated with post-traumatic stress syndrome and substance use disorders. Trotter et al.²⁷⁹ observed positive results in a 12- week trial of sertraline in an alcohol dependent sample of individuals (two-thirds female) with co-occurring post-traumatic stress syndrome. The results indicated significant reductions in post-traumatic stress syndrome symptomatology, substance abuse and symptoms of depression.

* Benzodiazepines for anxiety disorders can be contraindicated with substance dependent individuals because these drugs are especially susceptible to being abused and have a synergistic interaction with alcohol.

Co-occurring Severe and Persistent Mental Disorders

a) Prevalence and Etiological Issues

While many specific psychiatric disorders fall under the general rubric of "severe and persistent mental illness", the most common diagnostic categories are schizophrenia and bipolar illness. The literature on the prevalence of concurrent substance use and severe mental illness disorders is most recently summarized by Rosenthal and Westreich,³⁹ Drake and Muesser⁶⁴ and Mueser and colleagues.⁴⁴ The general conclusion from these reviews and others^4,40 is that between 40-60% of individuals with severe mental illness will develop a substance use disorder at some point during their lives, and about half currently meet criteria for substance abuse or dependence.^75,101 These rates are clearly higher than the prevalence of substance use disorders among people in the general population without concomitant psychiatric disorders. Diagnostic sub-groups within the general category of "severe and persistent mental illness" do not show a high preference for one type of substance over another. Alcohol is the most frequently abused followed by cannabis and cocaine.⁴⁴ Drug use preferences basically follow those in the general population.

Schizophrenia:

For schizophrenia specifically, population surveys have consistently shown elevated rates of alcohol use disorders (about three times the risk), and drug use disorders (about five times the risk).^75,76 Several studies reviewed by Mueser et al.⁴⁴ have also shown the high rates of substance use disorders among people with schizophrenia who are in treatment for the illness. Cuffel¹⁰¹ suggests that about half the youth at the first episode of schizophrenia present with or will develop a substance use disorder. Among people in treatment for alcohol abuse, the lifetime prevalence of schizophrenia is in the order of 4.5% to 6%, and among those in treatment with drug use disorders, 28%.

Bipolar disorders:

With respect to bipolar disorder, it is widely acknowledged to be the most common Axis I disorder to co-occur with substance use disorders. For example, in the classic ECA study by Regier et al.⁷⁵ the lifetime prevalence of any substance abuse or dependence among persons with any bipolar disorder was 56.1%. Further, the rates of substance abuse among those with bipolar disorder were several times higher than among those with unipolar depression. The high rate of co-morbidity has also been demonstrated among people in treatment for bipolar affective disorder, and among people seeking treatment for substance use disorders (see, for example, Weiss et al. ²⁸⁰ and Rosenthal and Westreich³⁹ for an overview of key studies and reviews).

As with all the concurrent disorders, co-occurring substance abuse/dependence and severe mental illness can interact in several complex ways that have important implications for screening, assessment and the planning of treatment and support. A detailed discussion of interaction issues specific to bipolar and substance use disorders and schizophrenia and substance use disorders is provided by Strakowski and Debbello²⁸¹ and Blanchard et al.²⁸² respectively. For severe mental illness generally, Mueser et al.⁴⁵ review the various etiological and interaction models and find considerable support for a super-sensitivity model whereby people with severe mental illness are more sensitive to the effects of alcohol and other drugs due to increased biological vulnerability and, therefore, experience more negative consequences from relatively small amounts of alcohol or other drugs

While more research is needed, experts generally agree that multiple factors are likely to be important for different groups of people, and even within the same person. It is generally understood, however, that severe and persistent mental disorders such as schizophrenia and bipolar disorder follow their course with or without significant improvement or recovery from the substance use disorder.

b) Screening and Assessment Issues

Screening for an alcohol or drug problem among people with concurrent severe and persistent mental illness can be accomplished with one or more of the screening tools identified in section 3.1.2. Indeed, it is this subgroup of people with whom most of the research on screening tools and assessment issues has been conducted. Specific examples for screening procedures and tools include the index of suspicion provided by Mueser and colleagues;⁴⁴ the DALI;¹⁷³ and the AUDIT.¹⁸⁷ Case manager ratings may also be particularly effective given the length of time that many people with severe mental illness may be involved with a case management program. The following assessment instruments have also been developed or tested specifically for this sub-group:

• the Alcohol Use Scale;²⁰⁶
• the Drug Use Scale;²⁰⁶
• the Substance Abuse Treatment Scale;⁶²
• Readiness-to-change.²¹⁸

An important implication of the super-sensitivity model discussed above⁴⁵ is that only a minority of people with concurrent substance abuse and severe persistent mental illness may be able to sustain controlled substance use. This is because moderate use may result in negative consequences or dramatically increase the risk of more severe substance use. It is critical to educate consumers about their biological sensitivity to the effects of alcohol and drugs. This process can begin during the assessment phase.

Assessing motivation and the stage of treatment is particularly critical as the findings are directly related to treatment planning.^62,218 Similarly, with respect to the assessment of medical and psychosocial needs (e.g., housing, food, health care). Mueser and colleagues⁴⁴ provide an up-to-date treatment of these, and many other assessment related issues and challenges with people with co-occurring substance abuse and severe and persistent mental illness. For this new sub-group it is recommended that the reader consult this new resource book from Mueser and colleagues.⁴⁴ See Appendix G for a list of common obstacles to assessment with this sub-group and potential solutions.

c) Treatment Implications and Clinical Issues:

Clients with combined severe mental illness and substance use disorders encounter serious cognitive and affective problems in addition to interpersonal difficulties. For instance, people in treatment for schizophrenia may experience hallucinations, have reduced emotional responses, and may at times be thought-disordered or express delusions. There is little harmony between cognition, emotional experiences and their expression. Since many of these individuals often have difficulty engaging in relationships with people in general, such may also be the case with a therapist. The decision-making process that is required to overcome inappropriate consumption patterns is affected by the ambivalence that characterizes these clients. Therapists that are working towards a remission of substance-related disorders may find it difficult to identify a real sense of motivation in their clients. Moreover, the misuse of substances will affect the pharmacodynamics of the medication taken for the primary disorder, exacerbating all the symptoms. Finally, the lack of social supports, adequate housing, meaningful daytime activity, as well as the functional impairment require a multifaceted approach to treatment and support.

d) Review of the Evidence

This section summarizes the evidence relevant to two questions:

• Is there evidence for or against a particular sequencing of interventions?
• Is there evidence that points to one or more specific treatment approaches or interventions as being particularly effective?

i) Sequencing:

There is a broad consensus, supported by research and current practice wisdom^35,44 that people with concurrent substance abuse and severe and persistent mental illness are best treated in an integrated program or system of services that deal concurrently with both the mental health and substance use problems.

ii) Intervention:

In addition to studies which have examined outcomes associated with the degree or types of program integrations³⁵ there have been a small number of studies of specific interventions for this sub-group. Jerrell and Ridgely²⁸³ conducted a partial experimental study (about 50 % of people were randomly assigned) to treatment conditions comparing a 12-step program, behavioral skills training and intensive assertive case management. Each of the latter two interventions was more effective than the 12-step condition on mental health outcomes and global life satisfaction. However, the effects on substance use were quite modest.

The effectiveness of the intensive case management and assertive outreach using Program for Assertive Community Treatment (PACT) model was examined in several of the studies reviewed by Drake et al.³⁵ This outpatient model involves a multidisciplinary team of specialized substance abuse and mental health professionals serving as the core resource and support team for a small number of consumers. It retains several features of the basic ACT model^284-286 including 24 hour, seven day a week coverage, assertive outreach, counseling and psychosocial supports complemented with specific substance abuse interventions tailored to the persons stage of change/stage of treatment motivation. In one of the better controlled studies,²⁰⁶ consumers in the PACT conditions showed more progress toward substance use recovery and decreased substance use severity.

Currently, the recommended compilations of empirically supported treatment and practice wisdom for this sub-group of people with concurrent disorders are the new resource book by Mueser and colleagues⁴⁴ and the recent review by Drake and Mueser,⁶⁴ Mueser and colleagues⁴⁴ first recommend the critical (i.e. minimal) components of a solid foundation of mental health services which must then be complemented by specialized concurrent disorders services.

The foundation of mental health services includes:

• medical management (including psychopharmacology);
• family support and education;
• supported employment;
• training in psychiatric self-management;
• crisis response services;
• housing;
• and, when needed, inpatient psychiatric hospital services.

The core components of specialized concurrent disorders treatment and support include:

• concurrent disorders assessment;
• clinical case management based on stages of treatment;
• motivational interviewing;
• a harm reduction approach (e.g. flexible goals);
• cognitive-behavioral substance abuse counseling;
• concurrent disorders group interventions, including social skills training groups;
• self-help liaison (e.g. Double Trouble; AA);
• work with families including behavioral family therapy and psychoeducation; and
• residential options, including housing.

Most consumers with co-occurring substance use disorders and severe and persistent mental illness will demonstrate serious impairment in several areas, such as work or school, family relations, judgment, thinking, or mood.²²⁵ As clinicians and support workers assisting this population know, these functional limitations will probably not disappear. The clients require detailed functional assessment, long term support, and an ongoing monitoring of functioning as part of the integrated treatment program.³⁵ In contrast with the heavy and regular use of substances by consumers with co-occurring mood/anxiety disorders, the consumption patterns of many of these consumers may have features more consistent with alcohol or drug abuse than severe dependence per se. This occurs, for example, when periods of heavy consumption are mediated by experiences of high stress.

While a high level of social-emotional support is likely to be needed to help the person with a severe mental illness remain in the community and remain at least minimally functional, the person's problems, risks and needs are greatly exacerbated by co-occurring substance use problems. Symptoms and related patterns of social functioning will normally be more negatively impacted by the interaction of two disorders. The fear experienced by many, especially after the first psychotic episode is quite overwhelming - sometimes characterized as "I thought I had lost control of myself and I would never come back." Social support often brings with it more stability and sufficiently adequate social functioning to remain in the community while taking part in an active and hopefully integrated treatment enterprise. With people with schizophrenia, for example, treatment and support, and the language of treatment and support, must be concrete, socially engaging and stabilizing. While self-help groups provide some support, people with severe mental illness must often rely on professional services for both therapy and social support. Active advanced case management that is able to integrate social support and some therapy holds promise.

As described above, a wide range of social supports may increase the likelihood of psychosocial stability, with sufficient integration and stability of the self to enable the person to actively participate in the ongoing assessment and treatment enterprise. These engagements are not likely to be consistent over time and mental health and substance abuse professionals must be prepared for instability of gains and regressions without blaming the client for being resistant to treatment, or consciously choosing relapse.

Co-occurring Substance Abuse and Personality Disorders

a) Prevalence and Etiological Issues

The results of a number of studies show that the overlap in personality disorders and substance use disorders is very high among people seeking help at substance abuse services.¹⁰⁴ Using DSM-III diagnostic criteria, the proportion of individuals attending substance abuse treatment that are found to have at least one personality disorder varies from 53 % to 100 %^102-104,295. The WHO classification system (ICD), which is stricter than the DSM criteria, yields more conservative estimates of the prevalence of personality disorders.^287,296 Individuals entering mental health services for the treatment of personality disorders also have high rates of concurrent substance use disorders.²⁸⁸

Substance use behaviour can lead to criminal activity, aggressiveness, disinhibition, chaotic lifestyle and therefore resemble personality disorder diagnostic criteria. Although overlap in the diagnostic criteria is apparent, studies that have examined this problem and adjusted for symptom overlap have still found substantial comorbidity.^288,289

Antisocial Personality Disorder:

While all types of personality disorders are found within individuals in substance abuse treatment, most research on the rates of co-occurrence concentrate on the cluster B disorders (e.g. antisocial, borderline, narcissistic). Antisocial personality disorder (ASPD) has received considerable attention because it presents the highest association with substance use disorders. Antisocial personality disorder is also the only Axis II personality disorder that can be assessed by lay interviewers using the Diagnostic Interview Schedule or DIS.^301,290 It is, therefore, the most commonly measured. The prevalence rates of antisocial personality disorder using DSM-III-R criteria in samples of people being treated for alcohol use disorder, vary from 10% to 53%.^{92,97,291,292,302}. In the Epidemiological Catchment Area Study³⁰³ the probability of presenting with a diagnosis of antisocial personality disorder was multiplied by four for men and by 12 for women, when an alcohol use disorder was present.⁹² Antisocial personality disorder, closely related to substance use disorders and conduct disorder, the childhood precursor to antisocial personality disorder, is a common antecedent problem among persons who subsequently develop mental illness. Further, antisocial personality disorder is more common among clients with severe mental illness than in the general population⁴⁵ Thus, the evidence is reasonably strong that antisocial personality disorder is a common etiological factor that may account for at least some of the higher rates of concurrent substance abuse among clients with severe mental illness.

Borderline Personality Disorder:

With the exception of individuals diagnosed as antisocial, those diagnosed with a borderline personality disorder (BPD) are more likely, compared to other subgroups of people receiving psychiatric services, to meet the criteria for a substance use disorder.^288,293-295 Studies on the co-morbidity of borderline personality disorder and substance use disorders report an overlap ranging from 15% -66%.^296-299 In studies of individuals attending alcohol treatment centres, reports of the prevalence of borderline personality disorder range from 13%-65%.^296,299,300 Kosten et al.¹³⁴ observed a rate of 12% of co-occurring borderline disorder in a sample of 150 people in treatment for opiate dependence. Epidemiological studies of individuals in the mental health system indicate even higher rates of co-occurrence between borderline personality and substance use disorders. In a study of 50 psychiatric outpatients who met criteria for borderline personality disorder, 84% had met criteria for a substance use disorder sometime in their life.²⁹⁵ In a Canadian study of people with borderline personality disorder, 23% met lifetime criteria for a substance use disorder.³⁰¹ A large study with a sample size of 2,463 examined the overlap between borderline personality disorder and substance use disorders among people admitted to a psychiatric hospital and found a prevalence rate of 21%.²⁹³

A number of issues should be kept in mind when interpreting these findings. These studies were almost exclusively carried out with individuals in treatment settings and, therefore, may be biased towards the detection of people experiencing greater dysfunction.

b) Screening and Assessment Issues

The validity of a diagnosis of personality disorder has long been a controversial subject.³⁰² Its onset in adolescence or early adulthood is a necessary condition for diagnosis. However, the reliable assessment of personality disorders, especially in individuals with substance use problems, poses significant challenges. Since the assessment of a personality disorder can be contaminated by the effects of the substance use, it can be difficult to separate out the effects of the drug use from behaviour that constitutes a true personality characteristic. This phenomena has been described as the trait-state artefact. Substances can increase mood instability, impulsivity, and interpersonal problems, which are features of personality disorders. Structured classification instruments and clinical interviewing are recommended to improve the reliability of diagnoses. A number of simple screening tools can be used to detect features of personality disorders. Self-reports such as the Personality Interview Questionnaire II³⁰³ or the Personality Diagnostic Questionnaire –4^th Edition³⁰⁴ provide a basic screen of Axis II symptoms. The Borderline Personality Disorder Scale³⁰⁵ is a useful approach to screen for symptoms associated with borderline personality disorder. The weakness with these self-administered questionnaires is that they have yielded high rates of false positives. These instruments should be utilized as a first step in a more comprehensive assessment.

Diagnosis of personality disorders should be conducted in a clinical interview with a qualified mental health practitioner (e.g., psychologist, psychiatrist, registered social worker). Structured interviews are recommended to facilitate more reliable diagnosis. The best and most widely known interview is the Structured Clinical Interview for the DSM-IV Axis II Personality disorders.³⁰⁶ Another well established interview is the International Personality Disorder Exam (IPDE)³⁰⁷ used by the World Health Organization. This instrument may not be as useful for clinicians since it is long and requires more training than the SCID. Both of these structured interviews for diagnosing personality disorders require clinical expertise in noticing clinically relevant criteria. In addition to assessing for the presence of a personality disorder, as with all individuals with concurrent disorders, an assessment of functioning is necessary. Those individuals with lower levels of functioning have a poorer treatment prognosis and will require more ongoing support.

It is important to assess the relationship between anger management difficulties and anti-social personality disorder. Many but not all people with an anger disorder will also have an anti-social personality disorder. Timing issues are important to investigate since resumption of substance abuse is a key predictor of relapse into an anger-aggression cycle.

c) Treatment Implications and Clinical Issues

The clinical implications of concurrent personality and substance use disorders have been described in a number of studies. Generally, individuals with both disorders experience greater dysfunction than their substance abuse counterparts without a personality disorder.²⁹⁷ Studies suggest that individuals with personality disorders not only have higher rates of substance use generally but are more likely to be polydrug users.¹³⁴ Further, a consistent finding is a higher comorbidity of personality disorders in drug users than alcohol users.³⁰⁸ People with a substance use and borderline personality disorder compared to substance abusers without borderline personality disorder are more likely to have a history of self harm behavior including suicide attempts; comorbidity for depressive disorders; and poor impulse control.^134,296,297 Many individuals with concurrent personality and substance use problems have high rates of chronic unemployment and lack social supports. These factors make it difficult for treatment to be effective since therapists cannot rely on a stable and constructive environment to support change.^309,310

Treatment of substance users who have a concurrent personality disorder poses a number of special challenges to clinicians. The clinical literature on individuals diagnosed with both disorders describes treatment as notoriously difficult. Interpersonally, these individuals are more rigid which contributes to greater difficulties in interpersonal relationships in the work place, with relatives and friends, as well as with therapists and treatment centres. These individuals often have greater difficulty with trust and intimacy. People with borderline personality who abuse alcohol and other drugs have been observed to be ambivalent about therapy and only moderately compliant.³¹¹ These individuals are typically more avoidant, especially of cues associated with negative affect, and this is associated with impulsive behaviour via rapid acting, mood-altering substances.³¹² For individuals with concurrent antisocial personality disorder, the ideas and behaviors that are at the core of the problem may be experienced as quite "natural". In addition, these individuals often have difficulty understanding the impact of their behavior on others. These features of the disorder increase the challenge for the therapist. Working with these individuals often leaves treatment providers feeling stressed. Consequently, it can be more difficult for clinicians to establish a working alliance, and retention in treatment is lower.³¹³

d) Review of the Evidence

This section summarizes the evidence relevant to two questions:

• Is there evidence for or against a particular sequencing of interventions?
• Is there evidence that points to one or more specific treatment approaches or interventions as being particularly effective?

i) Sequencing:

At this point much is unknown about what is the best approach to treating individuals with a concurrent personality and substance use disorder. The bulk of the literature to guide recommendations on treatment is primarily based on the results of research on the treatment of one disorder while tracking the impact on the other disorder. The results of these studies have conflicting implications for whether or not to utilize a concurrent treatment approach with this combination of disorders.

There is some literature, such as the recent study by Verheul et al.¹⁴⁵ to suggest that the best predictor of remission of the personality disorders is a recovery of the substance use. In other studies the components of 12-step programs within substance abuse programs have been observed to have an effect on personality structure.³¹⁴ Another study conducted a one-year follow-up of people attending a 6-8 week alcohol treatment program and found that people with concurrent borderline personality disorder showed significant improvements in leisure time satisfaction, decreased hospitalizations and stronger family relationships.³¹⁵ The implications of these research findings is that the focus of treatment should be the recovery of the substance use disorder, this being the key to the remission and/or improvement of the Axis II disorders.

In contrast to the above-mentioned studies, many other studies have found that the presence of a co-occurring personality and substance use disorder negatively affects the outcome of the index disorder. Kosten et al.¹³⁴ in a study of 150 opiate users observed that 2 ½ years following substance abuse treatment, individuals diagnosed with personality disorders had a poorer outcome. In other studies, the presence of antisocial personality disorder was associated with poorer social functioning, higher levels of substance use and worse response to traditional substance abuse treatment programs.¹³¹ The presence of antisocial personality disorder features has been linked to higher dropout rates from substance abuse treatment.³¹⁶ These findings suggest that the presence of the two disorders is associated with poorer treatment prognosis, and that the concurrent treatment of both disorders is critical. Many clinicians conceptualize the Axis II disorder as independent from the substance use disorder and that improvements in the area of substance use may be difficult to accomplish without simultaneously addressing the disturbances in character.

Studies of integrated models for treating co-occurring personality disorder and substance use disorders are scant. There is some recent evidence to suggest that treatment of borderline personality disorder and substance use disorders at the same time can be effective.³¹⁷ Further, prevailing clinical wisdom is that concurrent treatment is likely more sensitive to addressing the issues associated with borderline personality disorders and substance use disorders. In contrast to the treatment of clients with borderline personality disorder, most clinicians generally recommend routing clients with antisocial personality disorder and substance use disorders into substance abuse treatment. This trend likely reflects the lack of empirically supported treatments for antisocial personality disorder itself

ii) Interventions:

Though evidence for the specificity of psychosocial treatments has not been established, there is increasing interest and promise in cognitive behavioral approaches. Dialectical behavior therapy (DBT), developed by Linehan,^318,319 has recently gained considerable notice in the literature because of its empirical support. DBT is based on a biosocial theory that views the dysfunctional behavior as a problem-solving behaviour, which functions to soothe painful emotions.³¹⁸ From this perspective, while substance abuse behaviour is considered to be a maladaptive response, it is hard to change because it helps in the short run to modulate overwhelming, uncontrollable, and intensely negative emotions.³¹²

Linehan and Dimeff³¹² have developed a treatment manual that articulates an extension of standard DBT to fit the needs of borderline individuals with concurrent substance abuse disorders (DBT-S). The treatment emphasizes attachment strategies to increase retention, strategies to address urges and tendencies to use drugs, ad lib case management to provide coaching to address concrete needs, and pharmacotherapy where indicated with specific subtypes of substance users (e.g., methadone for opiate dependent individuals).

In a recent study, Linehan et al.,³¹⁷ randomly assigned 28 women to DBT or a treatment-as-usual condition of community-based care. The women who were diagnosed as having borderline personality disorder, were dependent on alcohol or other drugs and tended to have other psychiatric diagnoses. The women received one-year of treatment, including individual therapy and group skills training. Results indicated that those receiving DBT compared to the control group had significantly greater reductions in their substance use. The women in DBT were also more effectively retained in treatment; 64% compared to 27%. As well, those in DBT had better social and global adjustment after one-year of treatment, and at 16-month follow-up. A replication of this study is currently underway,³²⁰ and further studies with larger sample sizes are needed.

Regardless of the specific treatment approach, the degree of overall impairment should be used to guide treatment interventions. Some people with co-occurring personality and substance use disorders may not present with significant functioning difficulties. These individuals have the best prognosis. Others may present with moderate social and occupational impairment and lack coping skills. While one may expect an amelioration with a significant improvement or recovery from the substance use disorder, the prognosis is not as good as with higher functioning individuals. A frequent problem for individuals with co-occurring personality and substance use disorders is their serious impairment in several areas of functioning. Many of these individuals lack money, work, food and/or shelter. They may have been evicted from housing situations; be on no admit lists to housing resources; be living in abusive relationships; be engaged in very disturbed family relations; and may have trouble with the law. Assistance and ongoing support is needed for these problems. This can be achieved through intensive case management and ongoing monitoring of functioning. Assertive outreach may be necessary to reduce the likelihood of premature termination from treatment.³¹²

Co-occurring Substance Abuse and Eating Disorders

a) Prevalence and Etiological Issues

With increased concern about concurrent disorders in the past two decades, there has been more attention to the combination of substance use and eating disorders. Common types of eating disorders are anorexia nervosa, bulimia nervosa and binge eating. Most of the data on the co-occurrence of substance abuse and eating disorders is based on samples of people with the latter. While the association between eating and substance use disorders is higher than expected it also varies significantly across studies (3% to 49%).

Substance abuse is clearly higher among individuals with bulimia than those with restricting anorexia.^321-323 The reported rates of co-occurrence with anorexia nervosa are estimated to range from 6.7% to 23%.^324,325 In contrast, estimates of the prevalence of co-morbid bulimia nervosa and substance abuse problems range from 9% - 55%.^326-331 From the reverse perspective, studies of the prevalence rates of eating disorders within the substance abuse treatment population also yield highly variable estimates, ranging from 1% to 32%.^248,332-334 In one study of people with alcohol dependence, 10% had a lifetime history of anorexia nervosa and 20% had a lifetime history of bulimia.³³⁵In another study of individuals presenting to an alcohol treatment centre for the first time, 7% met current criteria for anorexia and 7% met criteria for bulimia.³³⁶ In a study of people abusing cocaine, 32% met DSM- III criteria for anorexia nervosa, bulimia or both, less than 1% of men and 4% of women met criteria for anorexia whereas 20% of men and 23% of women met criteria for bulimia.³⁴⁸

Divergent theories have been put forth to explain the association between eating disorders and addiction problems. These theories emphasize the biological, psychological and behavioral mechanisms that link them. One prevalent view is that both disorders reflect addictive disorders.^337-339 There is a growing opinion that eating disorders are different from other addictive behaviors. One theory is that the eating and substance use disorders are linked by underlying difficulties in the regulation of affect. From this perspective, the problematic behavior functions to regulate painful affect.^340,341

b) Screening and Assessment Issues:

Given the higher than expected prevalence of eating disorders in individuals with substance abuse problems, it is important for people to be routinely screened for the presence of an eating disorder when presenting to substance abuse treatment services. The importance of this initial screening must be underscored with clinicians working within the substance abuse treatment system who may be more inclined to overlook emaciation and poor appearance as a secondary effect of substance use. Inquiry should not be limited to individuals who appear underweight since there are higher rates of bulimia amongst individuals who abuse psychoactive substances. Individuals with bulimia may frequently appear of normal weight and, therefore, their symptoms may not be readily apparent.

Individuals with eating problems often do not volunteer information about their bulimia and compulsive eating without direct questioning about their symptoms and the problems may go undetected. A variety of assessment measures are available but tend to be lengthy or incomplete in their focus. The Psychiatric Screener discussed in the previous section on generic mental health screening is one tool that can be used to screen for an eating disorder.

More attention also needs to be directed at the detection of eating disorders in men. A growing number of studies^342,343 have observed high rates of eating disorders amongst men with substance use disorders. Since eating disorders are generally viewed as a disorder linked to women they are more likely to be overlooked in men. For example, clinicians have been observed to make more inquiries into eating disorder symptoms with thin women than with thin men.

For those people suspected as having an eating disorder on the basis of the initial screening, they should then be assessed using standardized diagnostic criteria (e.g., DSM-lV) to ensure reliable measurement. Assessment should focus on both current and past symptoms since clinicians caution that eating disorder symptoms often resurface as the substance use problems improve. Assessment of the eating disorder should also be repeated to determine its course in relation to possible improvement or worsening of the substance use disorder.

c) Treatment Implications and Clinical Issues

There are a range of physiological, physical, psychological and social consequences associated with the co-occurrence of eating disorders and substance use problems. Depending on the severity and chronicity of these disorders, their impact can range from mild to life threatening. Problems stemming from weight loss include fatigue, anxiety, sluggishness, amenorrhea and depression.³⁴⁴ Some individuals report using drugs, such as cocaine or methamphetamines which are anorectic agents, in the pursuit of low body weight. Others may rely upon substances to interrupt binges, elevate mood or to cope with the numbness. There is some evidence to indicate that eating and substance use disorders are associated with greater dysfunction, in particular problems with impulsivity, affective instability and instability. One study showed that individuals with concurrent anorexia nervosa and alcohol abuse had higher levels of theft, binge eating and purging.³²¹ Individuals with both disorders often have greater difficulty achieving abstinence and are more likely to relapse.³⁴² Other research indicates that substance using individuals with a concurrent eating disorder compared to those with only an addiction problem have:

• more severe substance problems (with the exception of heroin and tobacco);
• poorer health, more disturbed cognitive functioning (e.g., memory problems, confusion);
• greater affective instability;
• interpersonal problems; and
• poorer global functioning.³⁴³

The treatment of individuals with both disorders can be especially challenging to health professionals because few have the familiarity or expertise in the treatment of both disorders. The consequence of this is that clinicians often tend to focus on the treatment of the disorder with which they are most familiar. While expertise in both disorders is recommended, if this is not possible, consulting with other professionals who possess the requisite knowledge is necessary.

d) Review of the Evidence

This section summarizes the evidence relevant to two questions:

• Is there evidence for or against a particularly sequencing of interventions?
• Is there evidence that points to one or more specific treatment approaches or interventions as being particularly effective?

i) Sequencing:

The dearth of research literature on the treatment of both of these disorders limits the conclusions that can be drawn about the most effective treatment. There do not appear to be any controlled trials to evaluate the effectiveness of treating individuals with co-occurring eating and substance use disorders. This is largely due to the fact that until recently there have been no integrated treatment programs for people with concurrent substance use and eating disorders. Indeed, the presence of eating disorders in substance users has typically been ignored.

There are conflicting data on the effects of treating one disorder and tracking the other. A few studies have observed that the treatment of individuals with bulimia nervosa is not impacted by a prior history of psychoactive substance use.^345,346 On the other hand, other research shows that the treatment of one disorder without attention to the other can reduce the overall effectiveness of treatment.^336,347,348 Some clinicians speculate that as the substance use problems improve, the eating disorder may worsen due to symptom substitution. It may be difficult to break patterns of problematic eating and substance use without education and the development of effective coping strategies.

The prevailing clinical wisdom on the treatment of individuals with both disorders calls for interventions to be planned and implemented concurrently.³⁴⁹ Notwithstanding this view, there are compelling clinical reasons to suggest that if either of the disorders is so severe that it compromises the individuals' life, or critical aspects of functioning, treatment should first be targeted to that disorder.

Once the more serious disorder is stabilized the two disorders can be treated simultaneously. If neither of the disorders could be considered severe, or both are equally severe, they should be targeted simultaneously.

ii) Intervention:

Regardless of theoretical orientation, the majority of writers on this topic recommend a combination of:

• medical management to stabilize the individual;
• behavioral strategies to effect change in the eating and substance use behavior; and
• psychotherapy to address psychological issues.

A few integrated approaches to the treatment of these disorders have been described in the literature but are lacking empirical evaluation. A 12-step approach which integrates principles from Alcoholics Anonymous and Overeaters Anonymous has been described.³⁶¹ Within this model, both disorders are seen as being linked by an underlying addictive process. A few descriptions of inpatient programs for women with concurrent eating and substance use disorders are also described in the literature.^350,351 These programs integrate a variety of treatment strategies with common elements that include:

• a thorough psychosocial and medical assessment;
• monitoring of weight and food intake;
• nutritional assessment;
• dietary plans and supervision;
• behavioral modification to increase awareness of eating behaviors and stabilize aberrant behavior;
• education to address the denial associated with both disorders;
• psychopharmacology for severe symptomatology;
• individual and group therapy; and
• after-care.

At the Centre for Addiction and Mental Health, in Toronto, an integrated program for individuals with co-occurring eating disorders and addiction problems was recently established. The program provides outpatient treatment, based on a modified version of Linehan's³²⁹ Dialectical Behavioral Therapy model developed for individuals with borderline personality disorder.³¹⁸ The emphasis is on reducing relapse to eating and substance behaviors. Program participants attend weekly individual therapy and group skills sessions. Work is currently underway to evaluate this program.

Future research is needed to address what is the most effective treatment for this population.