Anxiety Disorders: Future Directions for Research and Treatment

Next | Previous | Table of contents

Chapter 4
Potential Directions for Future Research, Professional Care, and Professional and Public Education

The previous chapter provided a summary of the main research findings from a companion report, entitled Anxiety Disorders and Their Treatment: A Critical Review of the Evidence-Based Literature. This chapter provides a discussion of the gaps in the research literature, including methodological limitations, as well as of implications flowing from the research findings for professional care and professional and public education. It is hoped that the information presented in this chapter will provide a basis for discussions on effective treatment approaches for the anxiety disorders among the appropriate stakeholders in the health and mental health fields.

1. Gaps in the Research Literature

The quality of treatment research varies greatly across studies. One factor that may contribute to differences is the way in which different types of research are funded. Medication studies are often funded by pharmaceutical companies which allows for larger studies with more participants and greater resources. In contrast, studies of psychosocial treatments (including self-help treatments) are more frequently funded by public agencies or are conducted without external funding. In short, there are fewer resources available to researchers interested in cognitive and behavioural treatments, despite much evidence supporting their importance.

A number of gaps in the anxiety treatment research literature are discussed below. Methodological limitations in treatment studies for specific anxiety disorders are also highlighted.

i. Adequacy of Treatment Delivery

When reviewing the studies, it was not always clear whether investigators delivered treatments as reported. For example, it was possible that cognitive-behavioural therapists had different levels of skill. It is also possible that participants in the studies of pharmacological and psychological treatments were not compliant with treatment instructions. Additionally, most of the studies were not consistent in their measurement of treatment compliance and treatment integrity.

ii. Outcome and Other Measurement Issues

A limitation of nearly all treatment studies that were reviewed was their tendency to focus on symptom measurement only, at the expense of measuring functional impairment, quality of life, and other dimensions related to the impact of the disorder on the individual and family members. In addition, for some disorders (e.g., obsessive-compulsive disorder) the measures used in medication studies tended to be more sophisticated than those used in CBT studies, whereas for other disorders (e.g., generalized anxiety disorders), CBT researchers tended to use a broader range of measures than did pharmacotherapy researchers. For example, a study of the relative and combined effects of various cognitive and behavioural strategies for social phobia used “seeking additional treatment” as an indicator of treatment outcomes (Butler, Cullington, Munby, Amies, and Gelder, 1984).

In general, research studies of panic disorder with and without agoraphobia are associated with more sophisticated outcome measures than some other disorders (e.g., generalized anxiety disorders). PDA and PD studies have typically been more sophisticated in their designs and measures, including assessments of panic frequency, generalized anxiety, depression, agoraphobic avoidance and other domains of functioning.

Empirical evidence suggests that assessment of the complete impact of various treatment approaches should involve long-term follow-up. This was evident in a study comparing imipramine plus therapist-assisted exposure, imipramine plus therapist-assisted relaxation training, placebo plus therapist-assisted exposure, and placebo plus therapist-assisted relaxation for PDA. There were no differences between imipramine and placebo during treatment and through the one-year follow-up period (possibly due to the relatively low dosage of medication used). Therapist-assisted exposure led to significantly more improvement than relaxation, although differences were small (Marks et al., 1983). At two-year (Cohen, Monteiro, and Marks, 1984) and five-year follow-up (Lelliott, Marks, Monteiro, Tsakiris, and Noshirvani, 1987), participants continued to improve, although there were no longer differences among any of the groups.

iii. Assessment Instruments

Assessment instruments are used in both clinical and research settings to determine the presence or absence of symptoms of (in this case) anxiety disorders or to aid in clinical diagnosis (Health Canada, 1994). Many different instruments exist for each of the anxiety disorders, and agreement as to which ones are the “gold standards” for each specific disorder remains elusive. In addition, many of the instruments tap different domains; for example, some may measure psychological domains, whereas others may measure biological dimensions. The instruments may also vary in length, complexity (Health Canada, 1994) and psychometric properties (e.g., instrument reliability and validity) (Health Canada, 1996). As a result, comparisons between studies, even those that focus on the same anxiety disorder, are often difficult.

A related issue refers to whether the instruments are designed for use by (clinical) assessors or by patients. Patient assessments may result in different results than clinician assessments, as patients may assign more weight to certain domains being measured than clinicians. For example, a meta-analysis of antidepressants, behaviour therapy and cognitive therapy to treat obsessive-compulsive behaviour revealed that all forms of these treatments were more effective than placebo when based on assessor ratings. However, when comparisons were based on patient ratings, behaviour therapy and combined treatment tended to be more effective than antidepressants (van Balkom et al. 1994). Issues such as these need to be considered when conducting or reviewing treatment studies. (A list of useful references on assessment of anxiety disorders is included in Appendix 3 for the information of readers).

2. Methodological Limitations of Treatment Studies of
Specific Anxiety Disorders

i. Obsessive-Compulsive Disorder

In the case of OCD, many of the studies demonstrating the effectiveness of CBT have often been based on very small samples and have failed to use adequate controls. Although pharmacological studies of OCD have been better in this regard, they seldom use structured interviews to diagnose patients.

ii. Specific Phobia

Almost all studies of specific phobias have failed to use proper diagnostic criteria for identifying patients. In addition, although there are several studies examining treatments for animal and blood phobias, more controlled studies are needed for other phobia types (e.g., heights, claustrophobia, storms, flying, etc.). Finally, almost all behavioural studies for specific phobias have been based exclusively on exposure therapy.

iii. Generalized Anxiety Disorder

One difficulty with the GAD literature in particular is the fact that diagnostic criteria have changed quite dramatically over the years and most studies have relied on an outdated definition of GAD. When GAD was first introduced in DSM-III (American Psychiatric Association, 1987), it was conceptualized as a residual category for individuals with heightened anxiety lasting at least one month, who were not phobic, who did not meet criteria for panic disorder, and who were not depressed. With the publication of DSM-III-R (American Psychiatric Association, 1987), GAD was defined as a disorder in which the hallmark was excessive or unrealistic worry about two or more life spheres (e.g., work and family), lasting at least six months and accompanied by six of 18 associated symptoms. In DSM-IV (American Psychiatric Association, 1994), GAD is still a disorder of excessive worry lasting six months or more; however, the criteria have been revised, so that the worry must be difficult to control, be focused on a variety of topics (rather than two or more life spheres), and be associated with three out of six symptoms.

To date, most studies of GAD have been based on DSM-III criteria. Because of the revised criteria, it is likely that these older studies are no longer relevant to individuals meeting the current criteria for GAD. Therefore, the efficacy of pharmacological and psychological treatments for GAD, as the disorder is currently defined, has yet to be determined. Finally, outcome measures used in pharmacological studies of GAD have tended to be less sophisticated than those used in studies of psychological treatments.

iv. Social Phobia

For social phobia, studies have typically failed to differentiate between patients with generalized and discrete social phobias. Because of evidence that these two types of social phobias differ on a variety of dimensions, treatment studies should pay more attention to subtypes in social phobia research. For example, despite evidence that beta blockers (e.g., atenolol) are not helpful for generalized social phobia, they are often used in clinical practice to treat discrete social phobias (e.g., public speaking phobia). However, other than a few studies showing that beta blockers reduce anxiety in normal populations with heightened performance anxiety (e.g., musicians), there are no studies demonstrating their effectiveness in properly diagnosed patient populations. The use of beta blockers for performance anxiety should be investigated in patients diagnosed with discrete social phobias.

v. Posttraumatic Stress Disorder

For PTSD, very few treatment studies have been published and the few that are available have yielded inconsistent findings. Much work needs to be done in the area of developing and evaluating treatments for PTSD.

3. Potential Directions for Future Research

• Longitudinal research, using multidimensional approaches, is needed regarding risk factors for developing anxiety disorders. This is especially the case for disorders other than PD and PDA. In addition, there are virtually no studies that have examined the role of protective factors that might decrease the tendency to develop anxiety disorders among those considered to be at risk.
• Further research on comorbid conditions among persons with anxiety disorders is needed, particularly in light of the possible preventive implications of these conditions (e.g., social phobia).
• More controlled research, including meta-analytic studies, is needed on the relative and combined short- and long-term efficacy of pharmacological and psychological treatments for PTSD, specific phobias, social phobia and GAD.
• Methodologically-sound research on the effectiveness of other forms of psychotherapeutic approaches (e.g., psychodynamic and humanistic approaches) for the treatment of anxiety disorders is needed.
• Studies exploring treatment sequencing (i.e., the order in which different treatment components should be introduced) are needed in cases where combined treatments approaches are used.
• Long-term follow-up treatment studies are needed to explore possible differences in treatment efficacy over time (e.g., initial differences between treatments may wash out over time).
• More controlled research is needed to evaluate the effectiveness of newer SSRI's and other antidepressant medications in the treatment of the anxiety disorders.
• Treatment studies should include a broader range of outcome variables such as impact of anxiety disorders on quality of life, future health care utilization costs, lost wages, reduced productivity at work, and impact of treatment on families (including children).
• More data are needed on predictors of treatment response, as well as mechanisms by which treatments work, for all of the anxiety disorders. Once the efficacy of these treatments is established for different groups of patients, it will be important to find ways of predicting which treatments are likely to be effective for particular individuals, including those with one or more comorbid conditions, and to disseminate this information to clinicians and to the public.
• Virtually nothing is known about the effectiveness of treatment for the anxiety disorders by non-mental health professionals (e.g., family doctors). A variety of treatment manuals and training workshops have become available in the past few years, and it would be useful to assess the extent to which general practitioners can be trained to administer medications and CBT for anxiety disorders.
• Given the effectiveness of self-help (self-instruction) treatments and treatments involving minimal therapist contact for PD and PDA, it seems worthwhile to conduct more research on these approaches for other anxiety disorders.
• Controlled research studies as to the role and effectiveness of self-help/mutual aid approaches (e.g., participation in self-help groups) in helping individuals to cope with anxiety disorders should be undertaken. Preliminary research and anecdotal evidence suggest that many individuals (and their families) find participation in self-help groups beneficial.
• Although a critical review of measurement tools for the anxiety disorders was beyond the scope of this review, evaluation of these instruments is an important area for future research. A compendium and critical review of these instruments could be a useful first step to addressing this issue.

Because the state of the research varies for each of the anxiety disorders, some research recommendations may be identified which are specific to each type of disorder. These include:

PD and PDA:

• More research is needed on the effects of various forms of treatment in specific populations, including the elderly, children, culturally diverse groups, and individuals with multiple psychological problems (e.g., anxiety disorders and substance abuse).

OCD:

• Research on psychosocial interventions (e.g., exposure, response prevention, and cognitive therapy) is needed. More needs to be learned regarding the process of therapeutic change.
• Many of the older, uncontrolled studies should be repeated, using appropriate controls, adequate sample sizes, diagnosis using DSM-IV criteria (as measured by structured interviews), and adequate long-term follow-up.

Social phobia:

• Further research is needed to confirm preliminary research findings that CBT is at least as effective as pharmacological approaches in the short-term and probably more effective than medications in the long-term.
• The role of self-help approaches in social phobia remains to be studied.

GAD:

• Since relatively few studies are based on recent criteria, it is important for psychological and pharmacological treatments to be evaluated using properly diagnosed patients and a broad range of measures (including cognitive assessments).

Specific phobia:

• Studies that explore the efficacy of behaviour therapy with a broader range of diagnosed phobias (e.g., heights, storms, flying, et cetera) are needed.
• The efficacy of using strategies (e.g., medications, interoceptive exposure) shown to be effective for treating panic disorder for different specific phobia types remains to be investigated.

4. Other Implications flowing from the Review of the Evidence-
Based Anxiety Treatment Literature

• More education regarding the treatment of anxiety is needed for general health care professionals as well as for mental health care practitioners, including occupational therapists, social workers, psychiatric nurses and other clinicians. Unfortunately very little research has been conducted on training practitioners to treat anxiety. An exception is a study by Welkowitz et al. (1991) which showed that pharmaco- logically- oriented clinicians could be taught to deliver CBT to patients with PD.
• Research is needed on the most effective means of educating professionals about empirically validated anxiety treatment strategies. Incorporation of components on anxiety disorders and their treatment in training programs for psychiatric residents, family physicians, psychologists, occupational therapists, social workers, and other clinicians could be useful. However, a review of the effectiveness of educational programs in training professionals to deliver treatments for anxiety disorders should be undertaken as a first step.
• The development and dissemination of practice guidelines and structured assessments could help to facilitate the continuing education of health care professionals. In addition, incentives could be provided for professionals who treat anxiety disorders to seek additional training in CBT and other empirically validated approaches. Also, finding ways to make clinicians more accountable for the types of treatment they are using (e.g., periodic case reviews, making reimbursement contingent on using appropriate treatments) might encourage professionals to be better acquainted with current methods of treating anxiety disorders and other problems.
• Increased communication and linkages between general practitioners and mental health practitioners is needed. It has been suggested that specialized anxiety disorders clinics could be established to ensure that patients are offered the most up-to-date treatments for their problems. These clinics could also take a leadership role in training community-based health care practitioners to treat anxiety disorders. Other possibilities include the collaboration of community-based therapists with family physicians in the provision of psychological treatments in the community. Issues of cost-effectiveness and appropriateness of location of service delivery need to be explored.
• Improved communication between health/mental health practitioners and the self-help community and support group networks could contribute to enhanced knowledge and treatment of anxiety disorders. Funding experts to speak at support group meetings and funding training programs for individuals who lead support groups might improve the quality of self-help and support group programs available to individuals with anxiety disorders.
• More attention should be paid to educating the public about empirically validated treatments for anxiety disorders. Although there are several small organizations that hold support groups for individuals with anxiety disorders, little funding has been available to teach the public about anxiety disorders and where to get services in Canada. In contrast, the United States has a large national organization called the Anxiety Disorders Association of America (ADAA), whose membership includes patients and professionals with an interest in anxiety. This organization distributes a newsletter to members, provides referral information, and holds an annual conference to share new research findings with patients and professionals. A similar organization in Canada might help to educate the public about anxiety disorders and their treatment.
• Self-help (self-instruction) treatments are becoming increasingly viable options for individuals with anxiety disorders. Self-help manuals based on empirically validated treatments have now been published for PD and PDA, social phobia, OCD, generalized anxiety disorder, and specific phobia (Antony, Craske, and Barlow, 1995). In a time of shrinking health care resources, educating the public about empirically validated methods of self-help is an important objective.
• A number of other suggestions for improving public awareness include the development of a self-help or self-care handbook for Canadians with anxiety disorders, including coping strategies and resources available to Canadians. In addition, preparing fact sheets on each of the anxiety disorders might be an efficient way of disseminating information to general practitioners and to the general public.
• Finally, focusing on anxiety disorders as part of Mental Illness Awareness Week (an event co-sponsored by the Canadian Psychiatric Association) would help to increase awareness of the anxiety disorders and their treatment.

Next | Previous | Table of contents