The present study was motivated by the need to assess the effects of applying a group therapy programme to users of a public mental health facility who meet the clinical criteria for an anxiety disorder diagnosis. Anxiety disorders represent a large proportion of the cases referred by primary healthcare and GPs to the mental health sector. Data from the unit in which this study was carried out over 6 months in 2002 (Study of referrals to the Mental Health Unit [Unidad de Salud Mental – USM], unpublished) show that 33.75% of patients referred by primary healthcare staff to the USM were diagnosed with an anxiety disorder after assessment by psychiatrists or psychologists, and in accordance with the clinical criteria of the DSM-III-R.

The time saved to healthcare professionals and the general benefits of a group intervention have encouraged the formation of groups of patients with anxiety disorders for the application of a therapeutic strategy. The aim of the present study was to check the effectiveness of an intervention with patients who were being attended by our unit but did not show clear signs of improvement in the initial sessions of individual treatment. The choice of applied techniques employed in the sessions is based on several years’ experience of clinical intervention using cognitive-behavioural therapeutic methods with groups of patients diagnosed with anxiety at the USM, as well as on a review of new publications in the field. Before carrying out the present study, and with a view to updating knowledge in preparation for it, a final review of therapy proposals in manuals and bibliographic databases was carried out. The aim was to find out not only about the work that was being done, but also its results, so that they coult be compare with those that may be obtained in the present study. Below is a summary of the most relevant contributions found in this review.

The discussion of the topic in a manual from nearly 20 years ago (Vila, 1986) focuses on the Three-Dimensional Model of anxiety, and on the specific techniques to use according to the component affected: Systematic Exposure/ Desensitization (physiological component), Cognitive Restructuring/Stress Innoculation/Self-Instructions (cognitive component) or specific behavioural skills (behavioural component). Later manuals refer to diagnostic categories and treatment to be applied; this is the case of Echeburúa and de Corral (1991), who refer to the use of exposure techniques in Panic Disorder and Generalized Anxiety Disorders, suggesting positive expectations in exposure to anxiety-generating internal and external stimuli. Finally, specific packages of techniques for different disorders are described, such as protocols for the treatment of Panic from the Center for Stress and Anxiety Disorders, available as a manual by Craske, Rapee and Barlow (Craske & Lewin, 1997), the protocol designed by Botella and Ballester (1997) for Panic Disorder, or the proposal for multicomponent treatment of Generalized Anxiety by Dugas and Ladouceur (1997). In contrast to the case of the above treatments, when Deffenbacher (1997) proposed a Generalized Anxiety Management Training (Entrenamiento en el Manejo de la Ansiedad Generalizada – EMA), he suggested that its application in groups containing patients with other types of anxiety is not only possible but, indeed, beneficial. This suggestion is of interest in considerations, on the one hand, of the benefits of including patients with some variety of disorders in a single group, and on the other, of whether the techniques applied are so different as to make it necessary for groups to contain patients with a "single" diagnosis. This issue will be dealt with in a little more detail later in this article.

A search in the bibliographical database PsycLit from 1991 to 2000, together with another in Medline from 1999, selecting those studies in which the group administration of psychological techniques in anxiety disorders showed effective results and those that described application methods, revealed that the majority of articles referred to experiments involving a single disorder – Panic Disorder with or without agoraphobia. Indeed, of the articles selected for their relevance to the aims of this study, more than twice as many referred to single-disorder groups as to "mixed" groups.

In those abstracts that made reference to group intervention in patients suffering from different anxiety disorders, Panic Disorder, Social Phobia or Generalized Anxiety Disorder are those that are mentioned. The exception is that of a curious study presenting the results of applying a treatment package to heterogeneous groups of patients with psychiatric disorder (Manning, Hooke, Tannenbaum, Blythe et al., 1994). In this case, the authors confirmed the efficacy of a combined treatment of medication and cognitive-behavioural therapy through the scores on a series of inventories, finding statistically and clinically significant results in the applications and observations before and after the process. Other studies deal with the format of techniques used in therapy for Social Phobia or Panic Disorder (Barlow, 1992). Where there are other references to different anxiety disorders, they include: individual application of therapy in Generalized Anxiety and in Panic Disorder (Woodman, Noyes, Black, Schlosser & Yagla, 1999) with medication, with no change observed in Generalized Anxiety, but changes observed in Panic Disorder; or the case of application to a group of GPs in Germany (Schulze, Osen & Hand, 1997) of a short course of therapy to aid them in promoting self-help skills in patients with anxiety disorders, where improvement was noted in some cases (Agoraphobia, Simple Phobia and Social Phobia), but not in Panic Disorder or in Generalized Anxiety.

Most of the articles, however, deal with the application of cognitive-behavioural treatment, with or without medication, in patients with Panic Disorder alone. Some of these consider in the result how the treatment affects concomitant disorders (Neron, Lacroix & Chaput; 1995; Margraf, Barlow, Clark & Telch, 1993), referring to improvement in Generalized Anxiety and depressive symptoms with individual therapy and also with group therapies, and to the benefits of group therapy in Panic Disorder. The rest of the articles coincide in supporting the effectiveness of individual and group cognitive-behavioural therapy in the treatment of Panic Disorder, in comparison with a placebo (White, Keenan & Books, 1992), a control group (Telch, Lucas, Schmidt, Hanna et al., 1993; Petterson, & Cesare, 1996; Van den Hout, Arntz & Hoekstra, 1994) or a brief support psychotherapy group (Beck, Sokol, Clark, Berchick et al., 1992). Some researchers have considered the incorporation of some type of drug, finding that cognitive behavioural therapy is effective alone or in combination with fluvoxamine (Sharp, Power, Simpson, Swanson et al. 1996), or that combined with alprazolam it offers better results than with the drug alone (Bruce, Spiegel, Gregg & Nuzzarello, 1995).

In general, the procedures used are described under the heading cognitive behavioural therapy. Some works describe the components of these treatments, mentioning elements such as:

- Psychoeducation

- Some type of training in relaxation/breathing

- Cognitive therapy

- Exposure

As regards the exposure technique, some articles refer to its application in the face of feared ‘external’ situations (DiFilippo & Overholser, 1999) and others to interoceptive exposure (Telch et al., 1993); Barlow (1992) proposes its use with somatic and cognitive symptoms, while some authors propose the induction of symptoms (Rudd & Joiner, 1998).

As far as effectiveness is concerned, of the abstracts to the articles selected, only a few make quantitative references. The data indicate that 85% of those treated are free of anxiety crises, compared to 30% of controls (Telch et al. 1993); 63% of those treated maintain their improvement after 6 months (Beck et al. 1992). Absence of panic crises is found in 87% of the group that received cognitive therapy and 79% of those that received cognitive therapy after brief support psychotherapy. Botella and Ballester (1997), in a review on the topic, found that cognitive-behavioural programmes eliminate panic in 80% of patients, with improvements in 20-50% of the criteria (among them, use of health services).

A special mention should be reserved, finally, for the study by Echeburúa, Salaberría, Corral, Cenea and Berasategui (2000), given its similarity to the present work, with regard to place of application (Mental Health Centres), consideration of the medication among the data to be assessed, and sample with non-specific diagnosis (anxiety and depression). These authors reported a significant improvement in 70% of patients receiving cognitive-behavioural therapy with or without associated pharmacological treatment (according to patient’s profile), as against pharmacological treatment alone, for which no significant improvement was found.

The effectiveness of cognitive behavioural therapy would seem to be clear, as is the fact that group application is cost-effective. As the format is similar in the diverse anxiety treatments (relaxation, cognitive therapy, some types of exposure, etc.), it appears reasonable to think that a process adapted in the psychoeducational aspect to the different disorders (generalized, panic, agoraphobia, or even adaptive) where common techniques are applied may be a valid alternative. There remains the doubt over whether it would be effective, and, if so, how much so with respect to group therapy applied to patients with a single diagnosis. The improvement criteria may be diverse, but it would seem appropriate to include some clinical and some psychometric criteria, since the symptoms and their absence are difficult to define in such widely different diagnoses. In this study, one of the criteria chosen was that of clinical course: discharge from the USM would be the objective pursued, as this would signify clinical improvement (and relief of the burden on the unit’s resources). Continuity of the treatment would signify stagnation or worsening, and dropout would be difficult to define (search for other resources due to failure to obtain results, improvement and lack of interest in returning, etc.). Another criterion would be that of pharmacological treatment. Since the objective of this study is to track the patient’s progress and not to assess a drug, it seemed relevant to check whether the initial dose of the anxiolytic had been reduced, maintained or increased at the end of the group application or months later, if the pharmacological component is the same. Different components or their equivalent doseswere not assessed, on considering that a change of treatment is in itself a sign of therapeutic ‘failure’ of the previous one. In sum: when will the therapy in question be useful? When it reduces the volume of patients at the USM, permitting others to enter the therapeutic process; when reduction of the anxiolytic dose is achieved (or its elimination); if the user experiences some degree of subjective improvement expressed in a quantifiable way, and when the patient has had the opportunity to acquire forms of dealing with stressful situations or the anxiety itself.

Participants in this study were users of the Mental Health Unit (USM), located within a health centre belonging to Spain’s national health service, and who attended an anxiety management workshop during the period 1998-2000, inclusive. The basic criteria for attending the workshop were that the patient was receiving treatment at the USM (with a psychiatrist, a psychologist or both), and that the patient’s records showed, preferably, that they fulfilled the DSM-III-R criteria for an anxiety disorder (other, less precise diagnoses were considered as long as the predominant pathology was anxiety: adaptive processes, mixed anxiety-depression conditions difficult to classify, etc.). These patients were informed that they were to be involved in the study, subject to their formal agreement.

All the patients selected were called to an initial interview of an informative, motivational and clarifying nature, which also provided the opportunity for the first distribution of psychometric material; all those attending this first interview form part of the present study. Participants are divided into three groups according to their attendance at the different sessions; thus, Type 1 participants are those who attended the informative interview but did not come to a single session of the group workshop (at their own decision); Type 2 participants are those who began attending, but missed more than 50% of the sessions; and Type 3 participants are those who attended more than 50% of the sessions. In this way, the control group could be established on the basis of the user’s own decision to accept (or not) the application of the experimental procedure.

Table 1 shows the participants’ data. It should be pointed out that no statistically significant differences were found between the three types of user.

The present study included the assessment of some aspects that may be relevant for determining both the state of the sample and its clinical course at two different points: the beginning of the application of the psychoeducational process and the end of the group sessions.

In order to evaluate the stress to which participants had been subjected over the past year, Holmes’s "Inventory of Recent Experiences" (taken from Davis, McKay & Eshelman, 1986) was applied.

For differentiating the particular form of coping with stress, the "Questionnaire on Coping Behaviours in Stressful Situations", developed by Labrador (1992), after Muñoz (1988) was applied.

It was considered relevant to take into account a personality variable – alexithymia – that affects both the way stress is coped with and how far patients take advantage of the therapy. Alexithymia was measured using the Toronto Alexithymia Scale-20 (TAS-20), in its Spanish adaptation by Martínez Sánchez (1996), which has shown itself to be a psychometrically valid and reliable instrument (Bagby, Parker & Taylor, 1994; Bagby, Taylor & Parker, 1994).

A useful measure for anxiety was provided by Spielberger’s State-Trait Anxiety Inventory (STAI E/R), in its Spanish adaptation by TEA (1982), though in the present study only the ‘state’ variable was considered.

Another way of assessing anxiety was through the presence of tension indicated by the participant on Cautela’s Inventory of Tension and Anxiety Indicators (1977). A crucial aspect of this inventory is attention to the presence of physical signs indicative of anxiety; this differentiates it from the STAI, which does not consider such signs.

In order to determine the extent to which participants are affected by discouragement/depression, and feeling the need to use a tried and trusted instrument, we applied the Beck Depression Inventory (BDI), taken from Beck, Rush, Shaw and Emery (1983) was applied.

Also applied were some measures of general state that participants could indicate on a Likert scale of 1 to 10 were applied. These referred to their physical state (PHYSTA in this study) and psychological state (PSYSTA in this study). A list of 50 physical distress symptoms was provided so that participants could indicate the frequency with which they had experienced them in the previous week. Two types of measure were obtained: number of physical symptoms experienced (PHYSYM) and their mean frequency (FxPHY).

Finally, other types of subjective problem that may be reported by participants experiencing emotional distress in general ingeneral were considered, by means of a questionnaire taken from Davis et al. (1986), which includes a list of symptoms of specific response to stress, such as ‘low self-esteem’, ‘irritability’ or ‘resentment’, and on which participants judge the presence of these symptoms using a scale of 0 (no emotional distress) to 10 (extreme emotional distress). An average subjective intensity of distress score (SUBJ) was obtained from the quotient between intensity of the symptoms (0 to 10) and number of symptoms indicated.

In addition to the measures taken from the instruments mentioned above, information from the clinical records of those participating in the study was applied

. These included data such as "course of clinical care" (patients that had ceased to attend appointments, those that continued receiving treatment and those that had been discharged from the Unit) and "psychopharmacological treatment", referring to the prescription of benzodiazepines and their dose just before the beginning of the group sessions, at 3 months and at 1 year. The label TTO+ was used to refer to reduction of the dose (or its removal) by the patient’s psychiatrist, this being considered an indicator of clinical improvement; TTO- referred to maintenance or increase of the dose or change of drug, this being considered as a clinical indicator of no improvement.

The use of these indicators was intended to provide a more pragmatic way of assessing whether inclusion of participants in the group sessions and their attendance (or non-participation or poor attendance in the control cases) had any effect on clinical course. This, in turn, constitutes an assessment of whether application of this group therapy procedure is beneficial in relation to the relief of pressure on the USM’s care resources and the reduction of administration costs, in view of participants’ improvement level.

This group therapy consists of 8 weekly sessions lasting around 90 minutes each, with the same format in all three years in question (1998, 1999 and 2000) and application by the same therapist. The only difference between the three years was the involvement, mainly as observer but in some sessions as co-therapist, of a Psychology or Medical Resident working in the USM at the time. The sessions were predominantly psychoeducational, aimed, in addition to providing information, at helping users identify their problem and learn about different ways of managing their symptoms. Some time was always reserved for them to express their concerns and doubts and to receive support from the rest of the participants. It was suggested to them to apply between sessions the procedures dealt with in each one (their experiences in this regard could be discussed at the beginning of the next session). The essential content of each session was as follows:

First of all, the effectiveness of the group treatment`s application was determined through assessment of the clinical course of the users after attending the applied sessions by comparison with those subjects who did not attend them. Since the total of non-attenders is low, added to this ‘control’ group were the data from the Type 2 participants, that is, those who attended less than half of the sessions. Table 2 shows the distribution of users in accordance with the criteria of discharge, dropout (the user ceases to attend the USM) and treatment. The chi-squared statistic reveals that the differences observed in the table are significant. Although the percentage of experimental group participants who are discharged is higher than that for the control group (Type 1 and 2 participants), where the difference is most appreciable is in the dropout criterion; in fact, it is difficult to decide what indicates this difference: effectiveness of the treatment or participants’ adherence to the clinical process.

Since the period of care from a mental health unit like our own is prolonged beyond the clinical improvement itself due to the need for psychopharmacological treatment, clinical course was reviewed one year later in order to determine whether the initial data were maintained or had changed. To this end the clinical records were consulted once more, finding the figures shown in the table in Figure 1. This table reveals the marked greater likelihood of discharge and lower likelihood of dropout from the clinical process in those patients who received the group treatment.

The changes in anxiolytic drug prescription one year after attendance at the group sessions can be seen in Table 3, which indicates a statistically significant difference between participants: only 10.52% of "non-attenders" are on lower doses, as compared to 44.44% of "attenders". These data are included for their relevance with regard to practical (and also economic) aspects of the clinical course of those involved in this study, but it can be seen that the number of participants whose data is shown in this table is lower than that of the previous tables (given the difficulty of obtaining the necessary information), so that the mentioned difference must be claimed with a degree of caution.

Measurement instruments were also applied before and after the group sessions in which those participating in them reported on their psychological state and personal perception. Another way of determining the effectiveness of this group treatment is to observe whether the pre-post changes are significantly favourable. The data are shown in Table 4.

On carrying out the t-test for comparison of means, the most significant indicators were found to be those for state anxiety (STAI), depression (BDI) and general complaints of personal distress (desperation, bad mood, impotence; on a scale of 0 to 10), referred to as SUBJ.

To a lesser extent, significant values were found for reduction in the frequency with which participants felt physical distress (FxPHY) and subjective improvement in physical state (on a scale of 1 to 9). The other indicators, such as those for tension, number of physical problems perceived (PHYSYM) and assessment of psychological state (PSYSTA), also measured on a scale of 1 a 9, were not significant, but the means nevertheless indicate a positive result.

As already mentioned, the object of the present study was to assess the effectiveness of applying to a group of patients with anxiety disorders a set of sessions whose aim was to help them to know the characteristics of their disorder and offer them ways of managing anxiety. The form of determining the extent of the programme’s effectiveness involved measuring the variation in certain parameters in the control participants (Types 1 and 2 in this study), as compared to those to whom the group procedure was applied (participants who attended more than half of the sessions).

It should be mentioned that the participants’ characteristics do not appear to have influenced the results, as the differences between them were not statistically significant. This means that, having considered the variables sex, age, education, occupation, marital status and diagnosis, no decisive difference was appreciated between ‘experimental’ and ‘control’ participants. Also considered was a stable variable such as alexithymia (TAS-20), previous stress, duration of the care relationship for each patient, coping style and diagnosis, without any significant differences being discerned.

All the indications are that the results found in this study are not mediated by the individual variables mentioned. They would be expected, then, to indicate the extent to which they derive from patients’ having participated in the group sessions. A lack of significance was also found on ‘crossing’ these variables with the results for the dependent variables reported; this reinforces the previous statement.

The results reveal various aspects about what occurred with patients on attending the group:

If clinical course is considered as a relevant aspect, these data indicate figures for improvement close to those found in other studies mentioned in the Introduction. In some cases the figures in these other studies are quite similar, such as those of Echeburúa et al. (2000), with a mixture of anxiety and depression patients, which indicated an improvement in 70% of them. In the reviews of groups with homogeneous diagnosis, even though other improvement criteria are considered (absence of panic crises, for example), with figures of between 85% (Beck et al. 1992) and 80% (Botella & Ballester, 1997), when the reference is to the complete set of criteria, the ‘improvement’ ranges from 20% to 50% of the patients (including among these criteria ‘use of health services’) – results that are comparable to those reported in the present work.

In accordance with Deffenbacher (1997), the findings of the present work suggest that patients can benefit when they are mixed with those with other types of anxiety. The inclusion of patients with different anxiety disorders, especially Panic Disorder and Generalized Anxiety Disorder, involves the sole disadvantage of making it necessary to diversify the information at the psychoeducational points of the therapy; on the other hand, it provides a more comprehensive view of how anxiety affects people and how to cope with it (some coping techniques are the same in the two cases). Moreover, it is enlightening in the case of patients with comorbidity (the incidence of which seems to be considerable) and it aids professional care by saving time resources. Given the effectiveness data found in the present study, it would seem reasonable to continue to take advantage of the benefits of bringing together patients with different types of anxiety into a single group.

The failure to obtain data on all participants and on all the variables detracts from the power of the statistical results, and this is one of the weak points of the present study. In time, and with a more comprehensive body of data, these results will be more reliably defined, as will the possibility of some type of variable, not found to be relevant in this study, representing a factor of interest in subsequent revisions.

Thanks to Diego Torres, statistician at the Hospital General Universitario (Elche), for his invaluable help with the statistics, and to Dra. Mónica Godrid, who, during her time as a Resident specializing in Psychiatry, collaborated in the collection of data and made some pertinent suggestions in relation to data on psychoactive drug dosage.

References