The application of behavioral couples therapy to the assessment and treatment of agoraphobia

Clinical Psychology Review, Vol. 18, No. 6, pp. 663–687, 1998 Copyright  1998 Elsevier Science Ltd Printed in the USA. All rights reserved 0272-7358/98 $19.00 ⫹ .00

PII S0272-7358(98)00024-5

THE APPLICATION OF BEHAVIORAL COUPLES THERAPY TO THE ASSESSMENT AND TREATMENT OF AGORAPHOBIA: IMPLICATIONS OF EMPIRICAL RESEARCH Anthony D. Daiuto and Donald H. Baucom University of North Carolina at Chapel Hill

Norman Epstein and Samuel S. Dutton University of Maryland

ABSTRACT. The present review was undertaken to examine the literature regarding the role of the marital relationship in treatment of agoraphobia from a behavioral couples therapy (BCT) perspective. Both qualitative and quantitative analyses were conducted to evaluate (a) whether marital quality predicts treatment responsiveness and (b) how best to include the spouse in treatment of agoraphobia. Regarding the assessment of agoraphobics’ relationship quality, it is recommended that subjectively experienced ‘‘satisfaction’’ with the relationship be distinguished from more descriptive measures of general and agoraphobia-specific relationship ‘‘adjustment.’’ Treatment can include the partner as a ‘‘co-therapist’’ to assist with interventions directly targeted at the agoraphobia, or use BCT interventions to modify relationship interactions that might impede treatment gains. The application of BCT for these couples poses a number of challenges, including (a) the need to maintain a balanced treatment approach in an ‘‘unbalanced’’ system, (b) the need to target interventions at more than one treatment goal, and (c) the task of combining couple-focused interventions with those aimed at reducing phobic symptoms.  1998 Elsevier Science Ltd THE POTENTIAL role of interpersonal relationship factors in the etiology, maintenance, and treatment of agoraphobia frequently has been noted in the clinical literature, by writers with a variety of theoretical perspectives. For example, it has been proposed that an agoraphobic’s spouse benefits psychologically from his or her enhanced caretaking role and is likely to react negatively if therapy for the agoraphobic Correspondence should be addressed to Anthony D. Daiuto, University of North Carolina at Chapel Hill, Department of Psychology, CB #3270, Chapel Hill, NC 27599-3270; E-mail: [email protected]. 663

664

A. D. Daiuto et al.

partner leads to improved functioning and less dependence (e.g., Goodstein & Swift, 1977; Hand & Lamontagne, 1976). Bowlby (1973) theorized that agoraphobia develops from insecure attachment styles that become problematic adult working models for interpersonal relationships. Others have noted that conflicts over traditional sex roles within a marriage are resolved by agoraphobia (Hafner, 1986; Shean, 1990), as are ultimate stay/leave decisions for unassertive members of unsatisfactory marriages (Goldstein & Chambless, 1978). From a treatment perspective, those who have taken a cognitive-behavioral approach to intervention for agoraphobia (graded exposure, sometimes in combination with cognitive restructuring) often have enlisted the help of the spouse as a ‘‘cotherapist’’ based on the idea that the patient’s compliance with exposure homework assignments will be greater, and his or her general level of stress from marital conflict will be lower, when the patient is actively supported (and guided in the homework as needed) by the spouse (e.g., Barlow, O’Brien, & Last, 1984). The marital satisfaction of the nonagoraphobic spouse also is expected to increase (in contrast to the above psychodynamic view) as he or she gains empathy with the anxious partner’s experience and gradually is freed from the burdens of the caretaker role. Furthermore, because spouse-assisted exposure protocols tend to emphasize collaboration between the partners, this involvement may impact their relationship in positive ways (e.g., the nonagoraphobic partner increasingly encourages the spouse’s independence and cooperates less with avoidance behavior). However, there is no existing evidence documenting either positive or negative effects of spouse-assisted exposure on specific interaction patterns in couples’ overall relationships. After nearly four decades of speculation and debate about the role of the marital relationship in the treatment of agoraphobia, a number of recent reviews have attempted to summarize the empirical findings in this area. Following their metaanalytic review, Dewey and Hunsley (1990) concluded that (a) higher pretreatment relationship quality of the patient was associated with a greater reduction in agoraphobic symptomatology up to one year following treatment and (b) the effectiveness of ‘‘spouse-involved’’ exposure treatment was not significantly different from the effectiveness of ‘‘spouse noninvolved’’ exposure treatments. In a largely qualitative review that included limited quantitative analyses, Emmelkamp and Gerlsma (1994) concurred with the latter finding, but concluded that the effect of pretreatment marital quality on agoraphobic outcome is unclear. In their qualitative summary of the literature, Carter, Turovsky, and Barlow (1994) concluded that the quality of agoraphobics’ marital relationships predicts treatment outcome but, in contrast to the other reviewers, suggested that involving the partner in treatment may be more effective than treating the patient alone. The lack of conclusive findings regarding the impact of the marital relationship in treatment for agoraphobia has led some reviewers to suggest that the role of agoraphobics’ marital relationships should be de-emphasized in favor of a focus on the role of more broadly defined social support factors in treatment (Carter et al., 1994; Fokias & Tyler, 1995). However, despite the suggestion that a continued emphasis on the marital relationship is not likely to be fruitful, there is consensus that methodological and conceptual limitations have hampered attempts to elucidate the association between marital functioning and treatment for agoraphobia. For example, it has been noted in all of the above reviews that the field has not adequately conceptualized and assessed the quality of agoraphobics’ marital relationships. Similarly, although investigators have speculated that couples therapy for agoraphobia may be efficacious, there has been little agreement about how to include the partner in therapy. Thus, without a clear articulation of relevant relationship factors to guide them, most investigators

BCT and Agoraphobia

665

have simply used the spouses of agoraphobics as ‘‘co-therapists’’ in exposure therapy, without exploring the numerous ways in which the marital environment could potentially interfere with progress in treatment or be utilized to enhance outcome. Given these limitations, it seems worthwhile to consider that the field’s inconsistent and inconclusive empirical findings to date reflect a limited understanding of the marital functioning of agoraphobic patients that has obscured the need to explicitly address relationship issues as part of the treatment for agoraphobia. Clearly, there remains much to be learned about how the marital relationships of agoraphobics can be used to predict and influence their responsiveness to treatment. It is our belief that exploring the marital relationships of agoraphobics from a behavioral couples therapy framework will suggest guidelines for the field in terms of both assessment and treatment. Behavioral couples therapy (BCT) is by far the most widely evaluated approach to treating marital difficulties, and there is considerable evidence demonstrating its efficacy (Baucom, Shoham, Mueser, Daiuto, & Stickle, 1998). Briefly, BCT employs a skills-oriented format that emphasizes behavior exchange strategies and communication training (e.g., problem-solving, emotional expression). In recent years, greater attention has also been paid to modifying the specific maladaptive relationship cognitions that interfere with the effective use of these skills. It is important to note that BCT is not a single intervention strategy, but rather, an approach to treatment based on social learning principles. Taken together, BCT techniques have in common that they focus on the present, deal with behaviors and cognitions that are in the couple’s awareness, target specific interaction patterns to promote more adaptive functioning, and incorporate homework or the application of behavioral principles outside of the therapy sessions. The current review was undertaken to re-examine the literature from a BCT perspective and suggest a framework for better understanding how the relationship functioning of agoraphobics can be measured and treated. Specifically, two questions were evaluated: First, does pretreatment relationship quality predict agoraphobics’ responsiveness to treatment? Second, does inclusion of the partner in therapy improve treatment outcome? In each of the sections that address these questions, both metaanalytic data and qualitative analyses are used to (a) clarify conceptual issues raise by prior reviews, (b) apply a behavioral couples therapy (BCT) perspective to apparently conflicting findings, and (c) suggest a framework to guide future research in the field. Prior to addressing these issues, a few caveats regarding the scope of this review are warranted. First, as is consistent with the literature to date, the focus of this review is the marital relationships of agoraphobic patients; the role of nonprimary support relationships (e.g., friends, extended family members) has thus far received limited attention and will not be directly addressed here. Second, as virtually all of the treatment studies in this area have targeted white heterosexual female agoraphobics and their male partners, the generalizability of the findings and recommendations reported below is limited to patients with this demographic profile. Finally, although agoraphobic symptoms typically are the explicit target of treatment in the reviewed studies, it is important to note that all treated patients also experienced panic symptoms and that the reduction of subjectively experienced anxiety and panic is one of the field’s most frequently used outcome criteria. META-ANALYTIC METHOD Whereas many meta-analytic reviews are undertaken to evaluate hundreds of studies in a given field, a large number of studies is not essential for an informative meta-

666

A. D. Daiuto et al.

analysis. In fact, small scale meta-analyses can serve a critical function in guiding ongoing research in a field or stimulating interest in new or controversial treatment issues (e.g., Hahlweg & Markman, 1988; Hazelrigg, Cooper, & Borduin, 1987; ShohamSalomon & Rosenthal, 1987). Given the small number of studies in this area that were suitable for meta-analytic procedures, it must be emphasized that it is not a goal of this review to offer definitive statements about the empirical findings to date. Rather, it is our intention to suggest ways of approaching the existing literature and to offer suggestions regarding how this literature should be organized and evaluated in the future. The application of meta-analytic procedures provided a common metric upon which to compare the results of different studies, although these quantitative analyses also highlighted the great need for replication in this area.

Studies A computerized search of published and unpublished studies pertinent to the issues of marital quality and treatment of agoraphobia was conducted in two data bases, PsychLit and Dissertation Abstracts . The search covered the years 1977 through 1996. Combinations of the following key words were used to identify relevant studies: agoraphobia, treatment, outcome, marital, spouse, partner, and predictor. The bibliographies and articles located through these data bases provided a second source of studies. This search produced 62 published reports and 7 unpublished doctoral dissertations that were considered for inclusion in the meta-analysis. Reports meeting the following criteria were included in the meta-analysis: (a) subjects must have reported agoraphobic symptoms, and self-report measures of phobic complaints (i.e., subjectively experienced anxiety) must have been used; (b) studies had to have at least one group that received interpersonal skills training (e.g., communication, problem solving), or at least one group that received exposure treatment involving the partner, or individual exposure treatment in conjunction with measures of relationship quality; (c) statistics necessary for meta-analytic computations (e.g., means, standard deviations, or appropriate test statistics) must have been reported or made available through personal communication with the authors; and (d) to avoid the problem of nonindependence of reports, a single sample could not be represented more than once in any meta-analytic computation. When findings from a sample were reported in more than one article, those articles were considered as a single study. Relevant statistical information was extracted from this body of articles, but no statistic from a single sample was included more than once in a computation. Using these criteria, 20 independent studies qualified for inclusion in the meta-analytic procedures. The excluded reports consisted of book chapters, literature reviews, and clinical case studies, in addition to five empirical studies that did not address issues within the scope of this review.

Statistical Procedure The statistical procedures used in the present investigation have been extensively studied and written about by numerous authors (Cooper & Hedges, 1994; Hedges, 1981, 1982; Hedges & Becker, 1986; Hedges & Olkin, 1985; Ogles, Lambert, Weight, & Payne, 1990; Rosenthal, 1984; Smith, Glass, & Miller, 1980), and are briefly summarized below. Calculating effect sizes. The effect size (g) was defined as the difference between the mean scores of the ‘‘experimental’’ and ‘‘control’’ groups, divided by the pooled standard deviation. This formula is typically used when two different treatments or two

BCT and Agoraphobia

667

nonrandom groups receiving the same treatment are compared (Cooper & Hedges, 1994). When means or standard deviations were not reported, effect sizes were estimated from test statistics (e.g., t-test, F-ratio, Pearson r ) or significance levels. In some cases, when only means were available, standard deviations were obtained by substituting values estimated from other studies with comparable sample sizes and measures (Hedges & Becker, 1986). When the results for comparisons were reported as nonsignificant but data were not reported, a value of g ⫽ 0 was assumed. It should be noted that this procedure conservatively biases the results, as it is likely that the actual value of g would be slightly different from zero. Adjusting g for small samples. It has been shown that g gives a biased estimate of an underlying population effect when sample sizes are small. This bias was corrected by multiplying g by Hedges’s (1981) correction factor, c, which is a function of the size of the sample used in the study. The unbiased effect size estimator d, which incorporates this correction, was used in this investigation. Testing for heterogeneity. An assumption underlying the combination of a group of effect sizes is that the studies from which they were derived represent a ‘‘homogeneous’’ group that has been taken from a single population of studies. Thus, prior to estimating an overall treatment effect for a group of studies, a commonly used measure of the homogeneity of effect sizes, Ht, was used to determine whether the effect sizes from those studies varied systematically. This statistic is represented as the sum of squared differences between each effect size and their weighted average effect size. It has a chi-square distribution and measures whether the variability in the set of effect sizes is greater than would have been expected by sampling variability. A nonsignificant value of Ht indicated that the effect sizes (ds) of the included studies had a common population effect size. When Ht was significant, the studies were examined with the goal of identifying a subject or treatment characteristic that might moderate effect size. The studies were grouped based on this variable and Ht was partitioned into within-group (Hw) and between-group (Hb) homogeneity estimates that were separately tested for significance. When Hw was nonsignificant (each individual ‘‘subgroup’’ of studies was homogeneous) and Hb was significant (the subgroups were significantly different from each other), it was concluded that the hypothesized ‘‘moderator’’ variable had accounted for the heterogeneity among the entire group of studies, and an overall effect size estimator d . was computed separately for each of the homogeneous subgroups. Combining effect sizes across studies. After it was concluded that a given group of studies shared a common population effect size, the effect sizes (ds) were combined to estimate an overall effect size by computing a weighted average of the ds that took into account their respective variances. This weighted average, denoted by d ., is approximately normally distributed. The value of d . was tested for statistical significance against the null hypothesis, d .⫽ 0, by estimating its 95% confidence interval (p ⬍ .05). A significant overall treatment effect was obtained if the confidence interval did not include the value of zero.

Pretreatment Relationship Quality as a Predictor of Treatment Outcome Two of the reviews noted above (Carter et al., 1994; Dewey & Hunsley, 1990) concluded that the quality of agoraphobics’ marital relationships predicts their responsiveness to treatment (higher pretreatment relationship quality predicts a better out-

668

A. D. Daiuto et al.

come). These conclusions were based on the findings of studies that used measures of global relationship ‘‘satisfaction’’ to operationalize the construct of relationship ‘‘quality.’’ In contrast, Emmelkamp and Gerlsma (1994) found that, among the studies that predicted treatment responsiveness from relationship quality, effect sizes ranged from ⫺1.28 (the negative value indicating better results for distressed couples) to 1.59 (indicating better results for nondistressed couples). These results led them to conclude that the impact of relationship quality on treatment outcome is unclear. However, closer inspection of these data indicate that only 5 of the 20 effect sizes they reported are in the negative direction. Four of these come from two studies (Peter & Hand, 1988; Thomas-Peter, Jones, Sinnott, & Fordham, 1983) who used specific indices of marital functioning (e.g., expressed emotion, degree of dominance) other than relationship satisfaction. In other words, when pretreatment relationship quality is assessed solely in terms of the patient’s relationship ‘‘satisfaction,’’ the meta-analytic data of Emmelkamp and Gerlsma (1994) largely confirm the conclusions of the other reviewers. The conclusion reached by Emmelkamp and Gerlsma (1994) highlights the extent to which different operationalizations of the concept of ‘‘relationship quality’’ has obscured potentially important empirical findings. Most investigators have assessed the relationship functioning of agoraphobics with global measures of subjectively experienced relationship satisfaction. The most commonly used of these measures, for example, is the Maudsley Marital Questionnaire (MMQ; Crowe, 1978), which combines items regarding satisfaction with marital and sexual functioning with items that assess general satisfaction with work and social adjustment. Other measures used include the Marital Adjustment Test (MAT; Locke & Wallace, 1959) and the Dyadic Adjustment Scale (DAS; Spanier, 1976), both of which assess overall satisfaction with the relationship in combination with general items regarding disagreement and conflict between the partners. The use of such measures is problematic because most clinical and theoretical writers in this area do not emphasize the notion of relationship satisfaction as much as they focus on specific aspects of relationship functioning such as communication, control, dependency, closeness, and flexibility (Fry, 1962; Goldstein & Chambless, 1978; Goodstein & Swift, 1977; Holmes, 1982; Quadrio, 1984). In fact, many of these authors suggest that agoraphobic couples may actually appear quite dysfunctional in terms of these objective factors, while at the same time reporting relatively high levels of subjective satisfaction. The assessment of these couples requires a more complex definition of relationship quality. Within the marital field, numerous authors (Fincham & Bradbury, 1987; Glenn, 1990; Norton, 1983; Sabatelli, 1988) have argued that the construct of relationship ‘‘quality’’ is not tantamount to relationship ‘‘satisfaction.’’ In fact, these authors have asserted that two distinct constructs, ‘‘satisfaction’’ and ‘‘adjustment,’’ are both necessary components of relationship quality. In general, they concur that the former term, ‘‘satisfaction,’’ should be used to describe a partner’s global, subjective evaluation of the relationship, whereas the term ‘‘adjustment’’ is appropriate for referring to specific descriptive characteristics of the partners and their interaction (e.g., communication, problem-solving, cohesion). With this distinction in mind, we separately reviewed studies that have predicted treatment outcome from (a) global relationship measures that emphasized relationship satisfaction and (b) specific descriptive measures of how the partners interact with each other. Measures of relationship satisfaction. Seven studies reported data suitable for evaluating the influence of pretreatment relationship satisfaction on agoraphobic outcome. For posttreatment data, homogeneity analyses revealed a nonsignificant Ht ⫽ 9.35,

BCT and Agoraphobia

669

TABLE 1. Effect Sizes for Studies That Predicted Agoraphobic Outcome From Pretreatment Global Relationship Satisfaction Effect Size (d ) Study

Post

Milton & Hafner (1979) Emmelkamp (1980) Bland & Hallam (1981) Cobb, Mathews, Childs-Clarke, & Blowers (1984) Himadi, Cerny, Barlow, Cohen, & O’Brien (1986) Monteiro, Marks, & Ramm (1985) Arrindell, Emmelkamp, & Sanderman (1986) Peter & Hand (1988) d 95% confidence interval Upper limit Lower limit

— 0 1.79 0 0 .82 ⫺.12 0 .20 .49 ⫺.09

Follow-Up (Months) 1.69 (6) 0 (1) 2.02 (1–3) 0 (6) — .65 (7) .04 (3) — .38 a .79 ⫺.03

Note. Dashes indicate that data were not reported. d ⬎ 0 means that high satisfaction predicts greater improvement. d ⬍ 0 means that low satisfaction predicts greater improvement. a Does not include Bland and Hallam (1981).

χ 2 (6) ⫽ 12.59, ns, indicating that the effect sizes were from the same population of studies and could be combined to estimate an average treatment effect. As indicated in Table 1, the overall effect size suggested that patients with higher relationship satisfaction tend to have a better outcome, although this estimate was not statistically significant. For follow-up data, homogeneity analyses revealed a significant Ht ⫽ 11.72, χ 2 (5) ⫽ 11.07 p ⬍ .05, indicating that the effect sizes could not be combined to estimate an overall follow-up treatment effect. Closer inspection of these studies revealed that, whereas most of them relied solely on the patient’s self-report to determine ‘‘high’’ versus ‘‘low’’ relationship satisfaction before treatment, the study by Bland and Hallam (1981) divided subjects on the basis of (a) patient self-reported relationship satisfaction, (b) partner self-reported relationship satisfaction, and (c) clinicians’ ratings of an initial interview with the couple. Thus, this study, which yielded the greatest effect size at follow-up, operationalized relationship satisfaction more comprehensively than the others. Separating this study from the others resulted in a significant Hb ⫽ 4.97, χ 2 (1) ⫽ 3.84, p ⬍ .05, and a nonsignificant Hw ⫽ 6.75, χ 2 (4) ⫽ 9.49, ns, for the remaining five studies, which were combined to estimate an overall treatment effect. Table 1 presents the overall effect size for these follow-up studies, which is almost double that at posttest, but again is not statistically significant. In summary, at both posttreatment and follow-up, the overall effect size estimate indicated that agoraphobic patients with higher pretreatment relationship satisfaction tended to respond better to exposure therapy, although these results were not statistically significant. Measures of relationship adjustment. Whereas global measures of relationship satisfaction have most consistently been used to operationalize relationship quality, at least six different specific measures of relationship ‘‘adjustment’’ have been used to predict

670

A. D. Daiuto et al.

TABLE 2. Effect Sizes for Studies That Predicted Agoraphobic Outcome From Specific Indices of Relationship Adjustment Effect Size (d ) Study Thomas-Peter, Jones, Sinnott, & Fordham (1983)

Cohen (1986)

Peter & Hand (1988) Emmelkamp et al. (1992)

Measure

Post

Follow-Up

Management Effectiveness of Significant Other Rating Marital Patterns Test Affection given Affection received Dominance Couples Interaction Checklist Partner’s understanding of phobia Agreement about how to handle patient’s fears Degree of communication regarding the phobia Camberwell Family Interview Avoidance of Communication Scale

1.58

—

⫺.74 ⫺.60 .65

— — —

1.64 2.67

— —

2.5

—

0 ⫺.24

1.08 —

Note. Dashes indicate that data were not reported. d ⬎ 0 means that a high score on the measure predicts greater improvement. d ⬍ 0 means that a low score on the measure predicts greater improvement.

agoraphobic outcome following treatment. Table 2 presents these results, which were not combined into an overall treatment effect because the measures represent such a conceptually heterogeneous group. Results not amenable to meta-analytic calculations are discussed qualitatively. Thomas-Peter et al. (1983) conducted semi-structured interviews with the husbands of agoraphobics. Each husband was rated independently by two therapists on his ability to be supportive and successfully encourage and reinforce the patient’s independent activity, such that high scores on this Management Effectiveness of Significant Other Rating (MESOR) indicated that the husband had the potential to support the patient’s autonomy and independence, whereas low scores indicated that he was likely to accept and/or encourage the patient’s dependency. As indicated in Table 2, higher scores predicted greater agoraphobic improvement. Relationship satisfaction was not assessed in this study. Thomas-Peter et al. (1983) also used a self-report inventory called the Marital Patterns Test (Ryle, 1966) to predict outcome from more general marital interaction themes: Affection given by the patient, affection received by the patient, and the degree of the patient’s dominance in the relationship. Results suggested that a better outcome was achieved by patients who reported less affection both given and received, as well as a greater sense of dominance in their relationship. Peter and Hand (1988) used the Camberwell Family Interview (CFI) to classify couples into high and low expressed emotion (EE) groups based on the levels of criticism that the spouses expressed towards each other. Couples in the high EE condition were characterized by high levels of criticism, tension, disharmony, and dissatisfaction with the relationship (either open or covered), whereas those in the low EE condition had high levels of harmony, understanding, and warmth, as well as low levels of critical

BCT and Agoraphobia

671

attitudes. Although EE did not significantly predict posttreatment outcome, patients in high EE relationships tended to have better outcomes at 1- to 2-year follow-ups. The authors contrasted this finding with the fact that there was no significant association between outcome and either the DAS or the Marital Happiness Scale (Azrin, Naster, & Jones, 1973), both of which, in whole or in part, are regarded as measures of relationship satisfaction. Communication between the partners was assessed by Emmelkamp, Van Dyck, Bitter, Heins, Onstein, & Eisen (1992) with the Communication Questionnaire (CQ; Emmelkamp, Krol, & Sanderman, 1987), which contains scales that measure intimate communication, destructive communication, incongruent communication, and avoidance of communication. The couples in this study were nondistressed, as measured by the MMQ (Crowe, 1978), and their level of relationship satisfaction did not predict outcome. However, patients tended to have better outcomes when they had lower pretreatment scores on the avoidance of communication subscale of the CQ. Barlow and colleagues at the Center for Stress and Anxiety Disorders developed a self-report measure of couples communication, the Couples Interaction Checklist (CIC; Cohen, 1986; Craske, Burton, & Barlow, 1989), to assess communication specifically with regard to agoraphobia. Patients rated their partners, and partners rated themselves on five subscales: (a) partner’s understanding and knowledge of agoraphobia, (b) encouragement and support offered by the partner to enter feared situations, (c) pressure from partner to attempt phobic situations, (d) degree of communication with regard to the phobia, and (e) agreement between the partners about how to handle the patient’s fears. As indicated in Table 2, Cohen (1986) reported that three of these subscales significantly predicted treatment outcome, in spite of the fact that the seven couples in this study were largely nondistressed according to the DAS. Finally, Kitch (1983) used the Family Adaptability and Cohesion Scales (FACES; Olson, Sprenkle, & Russell, 1979) to predict the patient’s treatment response status from the couple’s pretreatment levels of cohesion and adaptability. Although limited data precluded the calculation of effect sizes, results indicated that couples in the ‘‘treatment responder’’ group were classified as ‘‘flexible’’ on the FACES adaptability dimension, whereas those in the ‘‘treatment nonresponder’’ condition were classified as ‘‘rigid.’’ Pretreatment levels of cohesion did not discriminate responders from nonresponders. All but one patient and one partner in this study of seven couples were maritally ‘‘satisfied’’ as measured by the DAS. Clearly, it is not possible to make definitive statements about the implications of these findings. None of these measures was used in more than one study, and many of the studies used very small samples. In addition, as psychometric data were not published about some of the measures, it is unclear how reliable they are or whether they actually measure the construct that they are designed to assess. Despite these limitations, three points are worth noting: First, whereas some of these measures assessed general characteristics of the relationship, others specifically focused on how the partners interact with regard to the agoraphobia. Second, in terms of the magnitudes of the effect sizes, many of these descriptive measures of relationship adjustment performed just as well as, if not better than, most measures of ‘‘satisfaction’’ as predictors of outcome in treatment studies. Third, many of these measures of adjustment predicted outcome even when the patients were satisfied with their relationships and level of satisfaction did not predict outcome. Thus, it appears that there is some empirical support for distinguishing between a subjective evaluation of relationship satisfaction and a descriptive analysis of relationship adjustment when assessing whether relationship quality predicts treatment outcome.

672

A. D. Daiuto et al.

Recommendations for Assessing the Relationship Quality of Agoraphobic Couples The results of this review suggest that a more in depth conceptualization of agoraphobic couples might be obtained by assessing them in terms of both ‘‘satisfaction’’ and ‘‘adjustment.’’ Specific recommendations for defining and measuring these constructs are presented below: Relationship satisfaction. Given the unreliable findings to date due to limited research, there remains a need to establish the extent to which relationship satisfaction, defined as an individual’s global subjective evaluation of his or her relationship, predicts the agoraphobic’s responsiveness to exposure therapy. This task is made difficult by the likelihood that the association between relationship satisfaction and agoraphobic symptoms is a highly complex interaction that may be more or less relevant depending on the couple. Even so, the utility of assessing this construct could be enhanced in a number of ways. For example, the field’s reliance primarily on measures of relationship satisfaction as an index of marital quality has been guided by the assumption that dissatisfaction with one’s relationship is a stressor that may have played an etiological role in the agoraphobia and will have a negative impact on treatment. In fact, several studies have demonstrated that relationship distress often precedes the onset of agoraphobia (Franklin & Andrews, 1989; Kleiner & Marshall, 1987; Last, Barlow, & O’Brien; 1984). It should be noted, however, that these studies also show that other stressors such as physical illness, childbirth, financial difficulties or nonspousal relationship difficulties may play a significant role in the onset of agoraphobia. Thus, it may be useful to assess the role of marital satisfaction in the context of a more comprehensive evaluation of life stressors that could potentially interfere with treatment or exacerbate agoraphobia. On the other hand, for many couples, relationship dissatisfaction may stem, at least in part, from the presence of the phobic symptoms. In such cases, relationship distress might be a good prognostic indicator (contrary to what has been predicted in the empirical literature), as it could motivate the partners to work together to eliminate the agoraphobic symptoms from an otherwise happy relationship. To date, empirical studies have failed to differentiate between relationship dissatisfaction as a contributor to agoraphobia and relationship dissatisfaction as a consequence of the disruptions to individual and couple functioning that are caused by the agoraphobic symptoms. Therefore, a more refined assessment of relationship satisfaction would involve asking the couple to rate the extent to which any dissatisfaction with the relationship is due to the presence of the agoraphobic symptoms. An additional strategy would be to establish the temporal association between the phobic symptoms and the relationship distress. Given that these approaches have proven fruitful for evaluating whether the nature of the causal association between relationship satisfaction and depression is relevant for predicting responsiveness to treatment for depression (O’Leary, Riso, & Beach, 1990), agoraphobia researchers should routinely incorporate them into their assessment of relationship functioning. Second, it should be noted that most investigators have used only the patient’s pretreatment relationship satisfaction to predict outcome. Consistent with clinical speculation that one or both of the partners may not acknowledge relationship difficulties, it may be that the combination of these sources of data or the differences among them predict outcome better than the patient’s absolute level of satisfaction. Although no single strategy for combining the data of both partners is routinely used, this issue has been previously addressed by cognitive-behavioral marital investigators (Baucom &

BCT and Agoraphobia

673

Mehlman, 1984; Jacobson, Follette, & Elwood, 1984), who have suggested approaches that are worthy of consideration. Baucom and Mehlman (1984), for example, attempted to predict marital status following BCT using additive, difference, sum plus difference, two gender, and ‘‘weak link’’ models for combining partners’ pretreatment data. Finally, as has been noted in prior reviews, it is critical that investigators choose measures of relationship satisfaction that are psychometrically sound. In addition, however, it is recommended that ‘‘pure’’ measures of satisfaction be used, rather than multi-component measures that typically include only a subset of items purported to measure the subjective evaluation of the relationship (e.g., MAT, DAS). The use of pure measures is critical if the construct of ‘‘satisfaction’’ is to be conceptually and empirically distinct from the set of variables that are used to measure the couple’s ‘‘adjustment.’’ For example, measures such as Snyder’s Global Distress Scale (GDS; Snyder, 1979) and Norton’s Quality of Marriage Index (QMI; Norton, 1983) are short, reliable indices of relationship satisfaction that do not overlap in item content with other measures of relationship ‘‘adjustment.’’ Indeed, some (Fincham & Bradbury, 1987) have argued that a single-item measure might be the most appropriate way to assess relationship satisfaction. General relationship adjustment. In contrast to ‘‘satisfaction,’’ the construct of ‘‘adjustment’’ refers to a wide variety of relationship phenomena that may relate to treatment outcome in different ways. Although limited, the empirical work to date suggests that various measures of relationship adjustment may prove useful in predicting treatment outcome. These are similar in that all provide descriptive information about how the relationship functions, as opposed to how the partners feel about the relationship. Taken together, these descriptive measures of general relationship adjustment indicate that patients who respond more to treatment may perceive that there is less affection in their relationship and feel a greater sense of dominance over their partners. Their relationships may be characterized by tension, disharmony, criticism, less avoidance of communication, and a flexible style of interaction. At first glance, some of these findings seem counterintuitive, as ‘‘healthy’’ couples are typically thought to be characterized by mutual emotional support and minimal conflict. However, the clinical and theoretical literature about agoraphobic couples consistently suggests that the spouses’ inability to acknowledge and address problems in the relationship can maintain agoraphobic symptoms and impedes treatment gains (Fry, 1962; Holmes, 1982). Viewed from this perspective, the apparently ‘‘unhealthy’’ relationship characteristics revealed by the above measures might actually indicate that the partners are able to achieve the level of autonomy necessary to tolerate and confront problems in their relationship and, thus, withstand the changes to the relationship that are inevitable if the patient is to benefit from treatment. This scenario could account for the recent finding of Schroeder, Frank, Funfstuck, Schneider, and Hahlweg (1996), who reported that agoraphobic patients who responded better to exposure therapy had reported higher levels of quarreling with their spouses before treatment. This speculation is consistent with the shift among BCT researchers toward a recognition of the adaptive value of conflict engagement in intimate relationships (Gottman, 1993; Markman, 1991; Sher & Weiss, 1991). The above interpretation of the data is offered as speculative, with the goal of encouraging future investigators to operationalize carefully the clinical observations and theoretical reports that originally motivated an interest in understanding agoraphobics’ relationships. Agoraphobic patients and their partners have consistently been

674

A. D. Daiuto et al.

described as mutually dependent, enmeshed yet emotionally detached, controlling, inflexible, uncommunicative, conflict avoidant, and sex-role stereotyped (Fry, 1962; Goldstein & Chambless, 1978; Goodstein & Swift, 1977; Holmes, 1982; Quadrio, 1984). These characteristics have been hypothesized to interfere with treatment by preventing the partners from accepting the changes in roles and interactional patterns that must occur to reinforce the agoraphobic’s treatment gains. In other words, there are specific aspects of the couple’s relationship ‘‘adjustment,’’ regardless of how satisfied the partners are, that are likely to influence how the patient responds to treatment. These factors have thus far received little attention in the empirical literature on agoraphobia. The BCT approach to assessing couple functioning offers a rich framework for delineating and assessing important relational phenomena. In particular, cognitivebehavioral researchers and clinicians have conceptualized couple functioning in terms of interrelated behaviors, cognitions, and emotions (Baucom & Epstein, 1990; O’Leary, 1987). The field has developed well-validated self-report measures and highly sophisticated observational coding systems for assessing problem-solving styles (Folkman & Lazarus, 1988), communication patterns (Christensen, 1987), and emotional expressiveness (King & Emmons, 1990). Despite the clinical and theoretical speculation (Goodstein & Swift, 1977; Holmes, 1982) that agoraphobic couples may function maladaptively in these areas, however, cognitive-behavioral assessment procedures have yet to be routinely employed to evaluate agoraphobic couples. In fact, only one agoraphobia outcome study to date (Craske et al., 1989) has attempted to examine the role of communication behaviors in treatment with BCT technology. Similarly, cognitive-behavioral marital researchers have identified a number of cognitive variables, such as partners’ attributions about each other’s behavior (cf. Bradbury & Fincham, 1990), that affect relationship adjustment, and these may be highly relevant to the assessment of couples experiencing agoraphobia. In addition to its traditional focus on identifying specific skill deficits and maladaptive cognitions, cognitive-behavioral researchers have recently begun to recognize the utility of assessing certain relational themes or dimensions thought to be central to the functioning of intimate relationships (Babcock & Jacobson, 1993). These include power (Babcock, Waltz, Jacobson, & Gottman, 1993; Gray-Little, Baucom, & Hamby, 1996), relationship investment (Baucom, Epstein, Daiuto, Carels, Rankin, & Burnett, 1996; Baucom, Epstein, Rankin, & Burnett, 1996), and closeness/intimacy (Baucom, Epstein, Daiuto, et al., 1996; Beach, Sandeen & O’Leary, 1990; Beach & Tesser, 1988; Christensen & Heavy, 1990; Prager, 1995). Not surprisingly, many of these themes have also been identified as salient for agoraphobic couples. For example, it has been suggested that a socialized imbalance of power in agoraphobic couples inhibits the spouses from resolving the long-standing emotional insecurities that predispose the wife to agoraphobic symptoms (Fodor, 1974; Quadrio, 1984). In addition, many clinical and theoretical writers have emphasized that agoraphobic patients and their spouses often fear separation and independence but are unable to achieve genuine emotional intimacy with each other (Fry, 1962; Holmes, 1982). Thus, it seems important that future investigators evaluate the roles of theoretically relevant relationship dimensions in the functioning of agoraphobic couples. Agoraphobia-specific relationship adjustment. Even though members of a couple might appear fairly well-adjusted in general, they may relate in a dysfunctional manner as a couple concerning the agoraphobia. For example, they might exhibit poor problem-

BCT and Agoraphobia

675

solving skills concerning ways to deal with agoraphobic symptoms or may fail to validate each other’s emotional reactions to the phobia. Craske et al. (1989) found that couples’ specific communication concerning agoraphobia (e.g., understanding and support by the nonagoraphobic partner) significantly predicted improvement in anxiety during exposure treatment, whereas global marital satisfaction did not. In terms of cognitions, partners may make particular attributions about the determinants of panic attacks and avoidance behaviors, which may differ from those they make about their overall relationship. Such specific assessment of behavioral and cognitive factors could differentiate a couple’s general relationship adjustment from those responses that are directly associated with the presence of agoraphobic symptoms in the relationship. Conclusions. The above recommendations are intended to generate new approaches to the assessment of agoraphobic couples that are grounded in the cognitive-behavioral marital literature. It is important to move beyond the assessment of partners’ general relationship satisfaction, to determine how their relationship can be characterized in terms of theoretically relevant relational adjustment variables that might impact treatment. A more sophisticated conceptualization of how the partners think, feel, and behave as a dyad should help not only to predict which patients are most likely to respond to treatment, but also to identify potential targets of intervention in the relationship and develop appropriate treatment techniques for addressing them.

The Impact of Spouse-Involved Therapy on Agoraphobic Outcome The second commonly asked question in this literature is whether the inclusion of the partner in treatment works better than treating the patient alone. Although interest in this issue stemmed from the view that agoraphobics may have problematic marital relationships, there has been little consensus about why involving the partner in treatment might be efficacious or how this should be accomplished. For example, some investigators have suggested that involving the partner in treatment will facilitate increased exposure practice for the patient ( Jannoun, Munby, Catalan, & Gelder, 1980); others have proposed that giving the couple a common goal to work on will indirectly reduce marital discord and enhance treatment outcome (Cobb, Mathews, ChildsClarke, & Blowers, 1984), and some have indicated that it is necessary to directly treat relationship problems that might underlie the agoraphobic symptoms and/or interfere with treatment gains (Arnow, Taylor, Agras, & Telch, 1985). Given the different rationales that agoraphobia researchers have used to justify including the partner in treatment, it is not surprising that the studies that have been conducted to evaluate the efficacy of partner-involved treatment have varied in a number of respects. Two important distinctions are worth noting: First, these studies have differed in terms of the target of the partner-involved interventions. Whereas many have focused exclusively on treating the phobic symptoms, others have also attempted to modify directly the functioning of the agoraphobic’s relationship. Second, these partner-involved treatments have utilized very different interventions (e.g., exposure, cognitive restructuring, communication training, problem-solving training) to achieve their desired goals. Thus, if research efforts are to facilitate a greater understanding of how the patient and partner can work together to ameliorate agoraphobic symptoms, the question of whether or not to include the partner in treatment should be considered in the context of other equally important questions about the target of the treatment and the specific intervention strategies used.

676

A. D. Daiuto et al.

Prior reviews maintained a focus on evaluating the ‘‘modality’’ (conjoint versus individual) of treatment for agoraphobia, resulting in conclusions that minimize the importance of attending to the marital relationship in treatment. However, it is necessary to note that most of the investigators in this area have attempted to ‘‘use’’ the relationship to treat the phobic symptoms by including the partner in exposure therapy. From the BCT perspective, this treatment approach seems counterintuitive, for two reasons. First, unless characteristics of couple relationships that can potentially interfere with exposure treatment also are addressed prior to the spouse-involved interventions, there is no a priori reason to assume that including the partner in exposure treatment will be helpful. In fact, a number of studies have indicated that members of couples that include an agoraphobic partner commonly report levels of marital conflict and distress (Arrindell & Emmelkamp, 1986; Dutton, 1997; Monteiro, Marks, & Ramm 1985), problem-solving deficits (Arrindell & Emmelkamp, 1986), emotional intimacy deficits (Dutton, 1997), and avoidant communication (Lange & van Dyck, 1992) that could impede the type of collaboration required in spouse-involved exposure protocols. Second, there is theoretical literature (e.g., Goldstein & Chambless, 1978; Hafner, 1986) and empirical evidence (e.g., Franklin & Andrews, 1989; Kleiner & Marshall, 1987) that relationship distress often precedes and may play an etiological role in agoraphobia. Thus, it is important not only to evaluate the impact of merely including the partner in exposure therapy; it is also necessary to evaluate whether direct treatment of relationship issues such as these leads to a decrease in agoraphobic symptoms. A number of questions about such interventions remain unanswered, including what specific intervention strategies should be used and how the treatment should be implemented (conjointly or individually). As such, we classified the outcome studies in this area into two broad types based on their focus of treatment: (a) those that directly target the phobic symptoms by involving the partner in exposure therapy and (b) those that used interpersonal skills training interventions to target the agoraphobic’s relationship functioning. In the sections that follow, each of these types of interventions is reviewed separately. Partner-assisted exposure therapy for the agoraphobic symptoms. Six studies have compared partner-assisted exposure with nonassisted exposure. Table 3 presents effect size data for these six studies. For posttreatment data, homogeneity analyses confirmed that the effect sizes were similar enough to combine into an overall treatment effect (Ht ⫽ 6.56; χ 2 (5) ⫽ 11.07, ns). The overall effect size for these studies indicated the superiority of nonassisted exposure, although this result was not statistically significant. As noted previously (Carter et al., 1994; Emmelkamp & Gerlsma, 1994), however, those partner-assisted exposure interventions that included training in cognitive coping skills (Cerny, Barlow, Craske, & Himadi, 1987; Woods, 1989) tended to be superior to nonassisted exposure treatments that also included this training. Three of these studies reported follow-up data with periods ranging from 1 to 6 months; these studies are also the ones that did not include cognitive coping training. As Table 4 indicates, the effect sizes from these studies were found to be homogeneous (Ht ⫽ .89; χ 2 (2) ⫽ 5.99, ns), and there was no significant overall difference between partner-assisted and nonassisted interventions. Cerny et al. (1987) reported data on 1- and 2-year follow-ups that supported the superiority of partner-assisted exposure. Again, this partner-assisted treatment included training in cognitive coping skills. Interpersonal skills training interventions targeted at relationship issues. Five studies have evaluated the efficacy of interpersonal skills training interventions targeted at relationship issues (either alone or in conjunction with exposure) relative to exposure treat-

BCT and Agoraphobia

677

TABLE 3. Effect Sizes for Studies That Compared Partner-Assisted Versus Non-Partner-Assisted Exposure Effect Size (d ) Follow-Up (Months)

Study

Post

Cobb, Mathews, Childs-Clarke, & Blowers (1984) Hand, Angenendt, Fischer, & Wilke (1986) Cerny, Barlow, Craske, & Himadi (1987)

⫺.13 ⫺.14 .32

Mannetter (1989) Woods (1989) Emmelkamp et al. (1992) d 95% confidence interval Upper limit Lower limit

.09 .48 ⫺.68 ⫺.19

.01 (6) ⫺.20 (6) .88 (12) .87 (24) — — ⫺.38 (1) ⫺.21 a

.12 ⫺.50

.12 ⫺.54

Note. Dashes indicate that data were not reported. d ⬎ 0 means that partner-assisted exposure improves more. d ⬍ 0 means that non-partner-assisted exposure improves more. a Does not include Cerney et al. (1987), which was the only study to report long-term followup data.

ment alone. Although the nature of these interventions is highly variable (e.g., individual vs. partner-involved, general vs. relationship-focused, implemented alone or in conjunction with exposure, etc.), all have in common an emphasis on teaching the patient how to function more adaptively in important interpersonal relationships. Thus, unlike exposure treatment, these interventions are targeted at the interpersonal difficulties thought to maintain agoraphobic symptoms and interfere with treatment (Goodstein & Swift, 1977). Table 4 presents a description of the treatments employed and effect size data for each of these studies. Despite their common interpersonal focus, homogeneity analyses revealed that these studies were not similar enough to justify averaging them into one combined effect size (Ht ⫽ 13.25; χ 2 (4) ⫽ 9.49, p ⬍ .05). Two hypotheses were considered to account for the variability among these studies. First, it was hypothesized that studies that employed interpersonal skills interventions in conjunction with exposure would be significantly different from studies that employed interpersonal skills interventions without exposure. This is based on the well documented efficacy of exposure for agoraphobia, and the expectation that providing only interpersonal skills training would not be efficacious in and of itself. Thus, studies 1 and 2 (skills training alone) should be different from studies 3, 4, and 5 (exposure plus skills training). However, partitioning the total heterogeneity into within and between groups statistics based on this criterion resulted in a significant value of Hw ⫽ 10.07, χ 2 (3) ⫽ 7.81, p ⬍ .05, and a nonsignificant value of Hb ⫽ 3.18, χ 2 (1) ⫽ 3.84, ns. These results indicated that the presence or absence of exposure in conjunction with skills training did not account for the variance among the five studies. It was then hypothesized that the studies would differ based on the target of the interpersonal skills training; in particular, whether it was specific to the marital relationship or broader in nature. Thus, studies 1, 3 ,and 5 (skills interventions targeted specifically at the couple) should be different from studies 2 and 4 (skills interventions

678

A. D. Daiuto et al.

TABLE 4. Effect Sizes for Studies That Compared Exposure Versus Interpersonal Skills Training Effect Size (d ) Study 1. Cobb, McDonald, Marks, & Stern (1980) a 2. Jannoun, Munby, Catalan, & Gelder (1980)

3. Arnow, Taylor, Agras, & Telch (1985)

4. Kleiner, Marshall, & Spevack (1987)

5. Crowe (1989)

Group 1 d 95% CI Upper limit Lower limit Group 2 d 95% CI Upper limit Lower limit

Treatments Compared

Post

Follow-Up (Months)

Partner-assisted exposure versus couples therapy directed at relationship problems Partner-assisted exposure versus partner-assisted problem solving directed at general life stresses and problems Partner-assisted exposure plus relaxation training versus partner-assisted exposure plus couples communication skills training Individual exposure versus individual exposure plus interpersonal problem-solving skills and assertiveness training Partner-assisted exposure versus partner-assisted exposure plus couples communication/problem-solving training directed at phobia-related conflicts

⫺1.25

—

⫺.77

⫺1.13 (4.5)

.80

.80 (8)

⫺1.05

⫺.62 (7.5)

⫺.05

.57 (3)

.15

.76

.77 ⫺.47

1.52 .04

⫺.90

⫺.87

⫺.33 ⫺1.47

⫺.30 ⫺1.43

Note. Dashes indicate that data were not reported. d ⬎ 0 means that interpersonal interventions plus exposure are better than exposure alone. d ⬍ 0 means that exposure alone is better than interpersonal interventions plus exposure. Group 1 includes Studies 1, 3, and 5. Group 2 includes Studies 2 and 4. CI ⫽ confidence interval. a Includes a mixed sample of agoraphobic and obsessive-compulsive patients.

targeted at general interpersonal problems). Indeed, partitioning these studies according to the focus of the skills interventions accounted significantly for the variation among them. The nonsignificant Hw ⫽ 7.18, χ 2 (3) ⫽ 7.81, ns, and significant Hb ⫽ 6.08, χ 2 (1) ⫽ 3.84, p ⬍ .05, indicated that skills interventions targeted at specific couple issues were different from general skills training techniques that emphasize assertiveness and problem-solving in a variety of interpersonal contexts. In short, the results supported the recognition of the marital relationship as a distinct factor in the treat-

BCT and Agoraphobia

679

ment of agoraphobia. This classification was replicated for the four studies that presented follow-up data. Whereas homogeneity analyses revealed a significant Ht ⫽ 12.27, χ 2 (3) ⫽ 7.81, p ⬍ .05, this variance was significantly accounted for by grouping them as above (Hb ⫽ 11.44, χ 2 (1) ⫽ 3.84, p ⬍ .05; Hw ⫽ .83, χ 2 (2) ⫽ 5.99, ns). Table 4 presents overall effect size data for the two subgroups of studies that emerged from these homogeneity analyses: Group 1 consists of the studies that compared exposure alone to exposure plus relationship-focused skills interventions. Although the treatments were not significantly different at posttest, exposure plus couple-focused interventions were significantly better at follow-up. This result was obtained even though the couples in the Arnow et al. (1985) and Crowe (1989) studies did not report relationship distress, suggesting that targeting relationship issues in therapy may be important even when the couple does not report relationship distress (couples in the Cobb, McDonald, Marks, and Stern [1980] study were maritally distressed). Group 2 includes studies that compared exposure alone to exposure plus general interpersonal skills training. At both posttest and follow-up, the former was significantly superior to the latter, indicating that the inclusion of more general interpersonal skills training interventions actually may have decreased the efficacy of the therapy. Taken together, the findings suggest that involving the partner in treatment may be most effective when the interventions employed are designed to produce changes in the partners’ interaction patterns. Whereas exposure therapy is likely the active therapeutic ingredient in terms of producing an immediate decrease in phobic symptoms, additional interventions aimed at fostering positive, direct, and constructive dialogues between the partners may serve to enhance the effectiveness of exposure by promoting the maintenance of treatment gains at follow-up.

Recommendations for Treating the Marital Relationships of Agoraphobics In conjunction with the assessment framework outlined in the preceding section, the empirical data above raise a number of questions about the different ways that the partners of agoraphobic patients may be involved in the treatment of agoraphobia. In particular, partners may be used to assist with the direct exposure treatment of the agoraphobia and/or become involved in addressing relationship issues that are deemed to be of relevance for treatment. Partner-assisted exposure. ‘‘Pure’’ partner-assisted exposure involves the significant other in treatment efforts that target the agoraphobia directly. These interventions utilize the partner as a ‘‘coach’’ or ‘‘co-therapist’’ to encourage the completion of homework assignments and reinforce the use of coping strategies taught by the therapist. Whereas some may apply only behavioral principles, others incorporate cognitive coping techniques that might enhance outcome. Although the relationship is not a direct focus of this treatment approach, a thorough understanding of couple functioning may be critical to implementing it effectively. For example, it is our impression that couples with healthy patterns of interaction (e.g., positive communication, mutual encouragement of autonomy, good problem-solving skills) may be the best candidates for partner-assisted exposure, whereas it might be countertherapeutic for those with poor relationship adjustment. This possibility suggests that it might be useful to assign couples to ‘‘partner-assisted’’ or ‘‘nonassisted’’ interventions based on their pretreatment level of relationship functioning. In the same vein, the efficacy of partner-

680

A. D. Daiuto et al.

assisted exposure and cognitive restructuring interventions for couples assessed as having relationship problems might be enhanced if they were implemented after relationship-focused interventions are employed to bolster the couple’s communication and problem-solving skills. In short, even though partner-assisted exposure does not directly target the relationship, a greater recognition of couple issues in treatment planning may enhance its efficacy. A second important issue regarding the use of partner-assisted exposure is the possible selection bias inherent in the fact that many of the couples who presented for therapy in the studies to date were nondistressed or only moderately distressed. In the presence of more severe relationship distress, partners may be less likely to agree to participate in treatment. It is for the agoraphobics in these couples that the question of utilizing nonspousal significant others may be most relevant. Couple-focused therapy. The second means of treating agoraphobia within a marital context is to focus directly on the marital relationship, with the goal of facilitating interactions between the partners that reinforce and perpetuate the patient’s attempts to overcome her fears. The results of the current review suggested that the use of couple-focused interventions for this purpose, in conjunction with traditional exposure interventions, may serve to enhance treatment outcome (compared to exposure implemented without additional couple-focused interventions). Such interventions are distinct from the ‘‘partner-assisted’’ approach that utilizes the partner as a cotherapist in the direct treatment of agoraphobic symptoms. BCT for agoraphobic couples can be used to address the general patterns of relationship functioning that are thought to characterize their interactions and thus contribute to overall stress for the agoraphobic (e.g., poor emotional expressiveness and problem-solving skills, power imbalances), as well as any agoraphobia-specific relationship issues that might potentially interfere with treatment responsiveness (e.g., how the partners communicate about the agoraphobia, individual lifestyle and relationship changes that may need occur once the agoraphobic partner begins to improve). This application of BCT raises a number of issues that should be considered if couple functioning is to be a target of treatment. First, it is important to emphasize that currently there are no data to suggest that couple interventions alone would be more effective than exposure therapy for most agoraphobic patients. As such, BCT is likely best utilized in combination with exposure. Indeed, BCT and exposure therapy complement each other in that both are skills-oriented, involve homework assignments, and require the therapist to be active and goal-directed. What remains unclear, however, is how the two treatments should be combined. Should they be implemented simultaneously or sequentially? If the latter, which (exposure or BCT) should come first? Although the literature to date has generally implemented couple and exposure treatments as separate sequential ‘‘components,’’ the simultaneous approach may prove most effective in terms of addressing the relationship issues that either interfere with, or are caused by, the patient’s initial treatment gains from exposure. This is an empirical question, however, that should be explored in future research. Second, it needs to be recognized that a primary goal of BCT for these couples is to facilitate a marital environment that reinforces the patient’s moves towards autonomy. Thus, the skills that are taught, and the way that the spouses use them, must support the goal of eliminating the agoraphobic symptoms. The application of BCT for this purpose may, at times, result in therapeutic recommendations for change that are initially quite disruptive and dissatisfying for the spouses. For example, despite mutual

BCT and Agoraphobia

681

satisfaction with an arrangement in which the husband does all of the household shopping, the therapist might ask the spouses to problem solve about changing the division of labor so that the wife does not remain at home as much. As such, decreases in relationship satisfaction may occur as the partners are encouraged to make changes in the ways that they interact. Such negative reactions to the loss of old ways of relating should be viewed as progress and normalized, as treatment-induced dissatisfaction with the relationship is likely to decrease the spouses’ motivation to work in a united way to eliminate the agoraphobia. Similarly, BCT can be used to offset some of the negative reactions spouses may have in response to the agoraphobic partner’s improvement (Hafner, 1977). For example, the use of cognitive restructuring techniques could be useful in modifying husbands’ negative attributions about their wives’ increased autonomy. Clearly, although reinforcement of the patient’s therapeutic gains is the primary goal of BCT for these couples, this is unlikely to be achieved unless continued attention is also given to the impact of treatment on the partners’ satisfaction levels. In addition, if a permanent dissatisfaction arises on the part of one or both partners, this needs to be recognized and the goals of the therapy need to be re-evaluated. A third issue to consider is the fact that BCT typically has focused on aspects of the marital relationship that are (a) in the partners’ awareness and (b) a source of current distress. To use BCT for the purposes described above, it will be necessary to direct the spouses’ attention to certain types of relational phenomena that they may not have previously noticed or acknowledged. This may seem unnecessary to those couples who do not report distress. Before skills techniques can be productively employed, then, it may be necessary to do more psychoeducational work with these couples than is typical for BCT so that they understand the reasons for focusing on their relationship. For example, a rationale for BCT that couples may find reasonable and persuasive is that attending to certain relationship patterns can serve an important preventive function, building their skills for any future stresses that may arise from treatment of the agoraphobia or from changes in their daily lives that will occur when the anxious spouse becomes relatively free of debilitating symptoms. A fourth issue for consideration is the fact that the BCT approach seeks to treat the relational system in a balanced way, encouraging the spouses to avoid blaming and focus on mutual responsibility for solving problems. In treating couples with an agoraphobic spouse, however, the therapist will be immediately confronted with an unbalanced system in which there is an identified patient. In fact, in addition to the agoraphobic symptoms, the ‘‘patient’’ is likely to have personality traits (e.g., a strong desire for affection, hypersensitivity to rejection, social withdrawal, a tendency to subordinate her own needs, mistrust) that significantly impact on the functioning of the relationship (Chambless, Renneberg, Goldstein, & Gracely, 1992; Chernen & Friedman, 1993). An advantage of BCT for these couples is that it can simultaneously address problematic dyadic interaction patterns between the spouses and the interpersonal schemata and overlearned behaviors that characterize each spouse’s attachment style. In other words, by focusing on the dyadic interaction in a balanced way, both the patient and partner will likely be required to confront and modify their individual interpersonal and coping styles. Thus, there may be a ‘‘synergistic’’ effect of treating dysfunctional and avoidant relationship functioning (in the areas of communication, problem-solving, emotional expressiveness) and avoidant agoraphobic behavior in the same treatment protocol. Finally, it is possible that the most distressed and poorly adjusted couples are less likely to seek conjoint treatment for agoraphobia. As such, it may be worthwhile to

682

A. D. Daiuto et al.

explore whether BCT could be useful in an individual, as opposed to a conjoint, format. In fact, Bennun (1997) has noted that relationship interventions with one partner can be a useful means of altering repetitive maladaptive interaction patterns and addressing presenting problems that are both interpersonal and individual in nature. Applied in an individual format, BCT could focus on helping the patient to understand dysfunctional behavior patterns in the relationship and identify strategies for altering them in the home environment, as well as modify relationship cognitions that contribute to distress. Although Bennun (1985) found that BCT interventions with individuals were as effective as those employed in conjoint and group formats for alleviating marital distress, additional research is needed to clarify whether some form of BCT could be useful in the absence of the partner in treating agoraphobia. For example, agoraphobic patients may benefit from BCT, even if their spouses do not participate in treatment, if it is implemented in a group format. This format would allow them to practice new relationship skills in a supportive environment. Conclusions. Despite the field’s primary emphasis on determining the efficacy of ‘‘spouse-assisted’’ therapy for agoraphobia, it is clear from the current review that the marital relationship has a number of important treatment implications. In fact, future research should be directed at evaluating the efficacy of interventions that directly target the marital relationship. Such interventions should be distinguished from ‘‘partner-assisted’’ interventions that use the spouse as a ‘‘co-therapist’’ to assist with the direct exposure treatment of the agoraphobia. ‘‘Couple-focused’’ interventions, as an adjunct to exposure therapy, may be employed to treat either ‘‘general’’ or ‘‘agoraphobia-specific’’ relationship issues thought to contribute to a stressful marital environment, perpetuate agoraphobic symptoms, or increase the patient’s risk of relapse. The goal-directed, skills-based approach of BCT is consistent with that of exposure therapy. Successful implementation of BCT for agoraphobic patients will require a thorough assessment of couple functioning, a convincing rationale to maintain both partners’ motivation to explore relationship issues, and a dual focus on maintaining the couple’s marital satisfaction and reinforcing the patient’s moves towards autonomy. It is hoped that future research will explore how best to combine BCT with exposure, as well as investigate how BCT might be adapted for severely distressed couples in which the partner is not willing to participate in treatment. FINAL REMARKS The present review was undertaken to highlight the ways in which the cognitivebehavioral marital literature might be used to guide both the assessment and treatment of couples in which one spouse is agoraphobic. Although the notion that the relationship functioning of agoraphobics may impact treatment has been controversial, it is important to note that this issue has been addressed by a relatively small number of empirical studies. As such, the meta-analytic results upon which many of our suggestions are based should be viewed as tentative, pending replication by future investigators. Despite this limitation, we hope that our approach to organizing and synthesizing the field’s findings to date will cast light on many of the unanswered theoretical and applied questions regarding the role of the marital relationship in treatment for agoraphobia. As was noted above, it is also necessary to point out that, in examining how relationship factors are related to treatment, both the theoretical and empirical literatures

BCT and Agoraphobia

683

focus almost exclusively on female agoraphobics. This emphasis reflects the well documented gender-skewed prevalence rates of agoraphobia; comprehensive epidemiological data indicate that 75.1% of agoraphobics are female (Wittchen & Essau, 1991). Nevertheless, 25% of agoraphobics are men, and the role of relationship factors for agoraphobic males may differ from their female counterparts. Thus, many of the conclusions and recommendations in this review may not generalize to male agoraphobics and their partners. Similarly, we know little about how the role of the marital relationship in treatment of agoraphobia is moderated by cultural and ethnic differences among these couples. In the OCD treatment literature, conflicting results have been obtained for partner/ family-assisted interventions that were employed in different cultural contexts (Emmelkamp, de Haan, & Hoodguin, 1990; Mehta, 1990) Although the treatments in these two studies were not completely identical, cultural differences between the treated couples may account for their discrepant results. It is recommended that such considerations be taken into account in the design and implementation of future studies with agoraphobic couples. Finally, although this review has focused on the ways in which the BCT literature might be used to guide future research and treatment in the area of agoraphobia, it is important to note that such collaboration should facilitate advancements in the theory and practice of BCT as well. For example, the challenge to better define and operationalize relationship quality for agoraphobic couples should serve to reinforce the BCT field’s efforts to more clearly define its central constructs and broaden its focus to include salient relationship themes. Similarly, whereas numerous methodological and statistical approaches have been used to combine and contrast the dependent data obtained from each of the spouses, BCT researchers have yet to reach consensus about how to define and measure the functioning of a ‘‘couple’’ as a single unit. Finally, the task of using BCT to foster a marital environment that supports the growth of an individual partner raises questions about how to identify the ‘‘client’’ and how to measure a ‘‘successful outcome.’’ All of these issues will need to be confronted if BCT is to remain a vital approach to studying and treating couples and realize its promise in the conceptualization and treatment of agoraphobia. Acknowledgment—This article was based, in part, on the dissertation of the first author, which was completed in partial fulfillment of the requirements for the doctoral degree in Psychology at the University of North Carolina under the direction of Donald H. Baucom. REFERENCES *References marked with an asterisk indicate studies included in the meta-analyses. *Arnow, B. A., Taylor, C. B., Agras, W. S., & Telch, M. J. (1985). Enhancing agoraphobia treatment outcome by changing couple communication patterns. Behavior Therapy, 16, 452–467. Arrindell, W., & Emmelkamp, P. M. G. (1986). Marital quality and general life adjustment in relation to treatment outcome in agoraphobia. Advances in Behavior Research and Therapy, 8, 139–185. *Arrindell, W. A., Emmelkamp, P. M. G., & Sanderman, R. (1986). Marital quality and general life adjustment in relation to treatment outcome in agoraphobia. Advances in Behavior Research and Therapy, 8, 139– 185. Azrin, N. H., Naster, B. J., & Jones, R. (1973). Reciprocity counseling: A rapid learning-based procedure for marital counseling. Behavioral Research and Therapy, 11, 365–382.

684

A. D. Daiuto et al.

Babcock, J. C., & Jacobson, N. S. (1993). A program of research on behavioral marital therapy: Hot spots and smoldering embers in marital therapy research. Journal of Social and Personal Relationships, 10, 119– 135. Babcock, J. C., Waltz, J., Jacobson, N. S., & Gottman, J. M. (1993). Power and violence: The relation between communication patterns, power discrepancies, and domestic violence [Special Section: Couples and couple therapy]. Journal of Consulting and Clinical Psychology, 61, 40–50. *Barlow, D. H., O’Brien, G. T., & Last, C. G. (1984). Couples treatment of agoraphobia. Behavior Therapy, 15, 41–58. Baucom, D. H., & Epstein, N. (1990). Cognitive behavioral marital therapy . New York: Brunner/Mazel. Baucom, D. H., Epstein, N., Daiuto, A. D., Carels, R. A., Rankin, L. A., & Burnett, C. K. (1996). Cognitions in marriage: The relationship between standards and attributions. Journal of Family Psychology, 10, 209– 222. Baucom, D. H., Epstein, N., Rankin, L. A., & Burnett, C. K. (1996). Assessing relationship standards: The Inventory of Specific Relationship Standards. Journal of Family Psychology, 10, 72–88. Baucom, D. H., & Mehlman, S. K. (1984). Predicting marital status following behavioral marital therapy: A comparison of models of marital relationships. In K. Hahlweg & N. S. Jacobson (Eds.), Marital interactions: Analysis and modification (pp. 89–104). New York: Guilford. Baucom, D. H., Shoham, V., Mueser, K. T., Daiuto, A. D., & Stickle, T. R. (1998). Empirically supported couple and family interventions for marital distress and adult mental health problems. Journal of Consulting and Clinical Psychology, 66, 53–88. Beach, S., Sandeen, E., & O’Leary, K. D. (1990). Depression in marriage: A model for etiology and treatment . New York: Guilford. Beach, S. R. H., & Tesser, A. (1988). Love in marriage: A cognitive account. In R. J. Sternberg & M. L. Barnes (Eds.), The psychology of love (pp. 330–358). New Haven, CT: Yale University Press. Bennun, I. (1985). Behavioral marital therapy: An outcome evaluation of conjoint, group, and one spouse treatment. Scandinavian Journal of Behavior Therapy, 14, 157–168. Bennun, I. (1997). Relationship interventions with one partner. In W. K. Halford & H. J. Markman (Eds.), Clinical handbook of marriage and couples interventions (pp. 451–470). Chichester, UK: Wiley. *Bland, K., & Hallam, R. S. (1981). Relationship between response to graded exposure and marital satisfaction in agoraphobics. Behaviour Research and Therapy, 19, 335–338. Bowlby, J. (1973). Attachment and loss (Vol. 2), Separation: Anxiety and anger. New York: Basic Books. Bradbury, T. N., & Fincham, F. D. (1990). Attributions in marriage: Review and critique. Psychological Bulletin, 107, 3–33. Carter, M. M., Turovsky, J., & Barlow, D. H. (1994). Interpersonal relationships in panic disorder with agoraphobia: A review of empirical evidence. Clinical Psychology: Science and Practice, 1, 25–34. *Cerny, J. A., Barlow, D. A., Craske, M. G., & Himadi, W. G. (1987). Couples treatment of agoraphobia: A two-year follow-up. Behavior Therapy, 18, 401–415. Chambless, D., Renneberg, B., Goldstein, A., & Gracely, E. (1992). MCMI-diagnosed personality disorders among agoraphobic outpatients: Prevalence and relationship to severity and treatment outcome. Journal of Anxiety Disorders, 6, 193–211. Chernen, L., & Friedman, S. (1993). Treating the personality-disordered agoraphobic patient with individual and marital therapy: A multiple replication study. Journal of Anxiety Disorders, 7, 163–177. Christensen, A. (1987). Detection of conflict patterns in couples. In K. Hahlweg & M. J. Goldstein (Eds.), Understanding major mental disorder: The contribution of family interaction research. New York: Family Process Press. Christensen, A., & Heavy, C. L. (1990). Gender and social structure in the demand/withdraw pattern of marital conflict. Journal of Personality and Social Psychology, 59, 73–81. *Cobb, J. P., Mathews, A. M., Childs-Clarke, A., & Blowers, C. M. (1984). The spouse as co-therapist in the treatment of agoraphobia. British Journal of Psychiatry, 144, 282–287. *Cobb, J. P., McDonald, R., Marks, I., & Stern, R. (1980). Marital versus exposure therapy: Psychological treatments of co-existing marital and phobic-obsessive problems. Behavioural Analysis and Modification, 4, 3–17. *Cohen, S. L. (1986). Marital communication skills training with agoraphobics . Unpublished doctoral dissertation, State University of New York at Albany. Cooper, H., & Hedges, L. V. (Eds.). (1994). The handbook of research synthesis . New York: Russell Sage Foundation. *Craske, M. G., Burton, T., & Barlow, D. H. (1989). Relationships among measures of communication, marital satisfaction, and exposure during couples treatment of agoraphobia. Behavior Research and Therapy, 27, 131–140.

BCT and Agoraphobia

685

Crowe, M. J. (1978). Conjoint marital therapy: A controlled outcome study. Psychological Medicine, 8, 623– 636. *Crowe, C. M. (1989). The effectiveness of problem-solving and communication skills training in enhancing treatment outcome for agoraphobic females . Unpublished doctoral dissertation, University of South Carolina, Columbia. Dewey, D., & Hunsley, J. (1990). The effects of marital adjustment and spouse involvement on the behavioral treatment of agoraphobia: A meta-analytic review. Anxiety Research, 2, 69–83. Dutton, S. S. (1997). Agoraphobia and interpersonal relationships . Unpublished master’s thesis, University of Maryland at College Park. *Emmelkamp, P. M. G. (1980). Agoraphobics’ interpersonal problems: Their role in the effects of exposure in vivo therapy. Archives of General Psychiatry, 37, 1303–1306. Emmelkamp, P. M. G., de Haan, E., & Hoodguin, C. A. L. (1990). Marital adjustment and obsessive compulsive disorder. British Journal of Psychiatry, 156, 55–60. Emmelkamp, P. M. G., & Gerlsma, C. (1994). Marital functioning and the anxiety disorders. Behavior Therapy, 25, 407–429. Emmelkamp, P. M. G., Krol, B., & Sanderman, R. (1987). The assessment of relationship beliefs. Personality and Individual Differences, 8, 775–780. *Emmelkamp, P. M. G., Van Dyck, R., Bitter, M., Heins, R., Onstein, E. J., & Eisen, B. (1992). Spouse-aided therapy with agoraphobics. British Journal of Psychiatry, 160, 51–56. Fincham, F. D., & Bradbury, T. N. (1987). The assessment of relationship quality: A reevaluation. Journal of Marriage and the Family, 49, 797–809. Fodor, I. G. (1974). The phobic syndrome in women: Implications for treatment. In V. Franks & V. Burtle (Eds.), Women in therapy: New psychotherapies for a changing society (pp. 132–168). New York: Brunner/ Mazel. Fokias, D., & Tyler, P. (1995). Social support and agoraphobia: A review. Clinical Psychology Review, 15, 347– 366. Folkman, S., & Lazarus, R. (1988). Coping as a mediator of emotion. Journal of Personality and Social Psychology, 54, 466–475. Franklin, J., & Andrews, G. (1989). Stress and the onset of agoraphobia. Australian Psychologist, 24, 203– 219. Fry, W. F. (1962). The marital context of an anxiety syndrome. Family Process, 1, 245–252. Glenn, N. D. (1990). Quantitative research on marital quality in the 1980s: A critical review. Journal of Marriage and the Family, 52, 818–831. Goldstein, A., & Chambless, D. (1978). A reanalysis of agoraphobia. Behavior Therapy, 9, 47–59. Goodstein, R. K., & Swift, K. (1977). Psychotherapy with phobic patients: The marriage relationship as the source of symptoms and focus of treatment. American Journal of Psychotherapy, 31, 284–293. Gottman, J. M. (1993). The roles of conflict engagement, escalation, and avoidance in marital interaction: A longitudinal view of five types of couples [Special Section: Couples and couple therapy]. Journal of Consulting and Clinical Psychology, 61, 6–15. Gray-Little, B., Baucom, D. H., & Hamby, S. L. (1996). Marital power, marital adjustment, and therapy outcome. Journal of Family Psychology, 10, 292–303. Hafner, J. (1986). Marriage and mental illness: A sex roles perspective . New York: Guilford. Hafner, R. (1977). The husbands of agoraphobic women: Assortive mating or pathogenic interaction? British Journal of Psychiatry, 130, 233–239. Hahlweg, K., & Markman, H. J. (1988). Effectiveness of behavioral marital therapy: Empirical status of behavioral techniques in preventing and alleviating marital distress. Journal of Consulting and Clinical Psychology, 56, 440–447. *Hand, I., Angenendt, J., Fischer, M., & Wilke, C. (1986). Exposure in-vivo with panic management for agoraphobia: Treatment rationale and longterm outcome. In I. Hand & H.-U. Wittchen (Eds.), Panic and phobias: Empirical evidence of theoretical models and long-term effects of behavioral treatments (pp. 104–127). Berlin: Springer-Verlag. Hand, I., & Lamontagne, Y. (1976). The exacerbation of interpersonal problems after rapid phobia removal. Psychotherapy: Theory, Research, and Practice, 13, 405–411. Hazelrigg, M. D., Cooper, H. M., & Borduin, C. M. (1987). Evaluating the effectiveness of family therapies: An integrative review and analysis. Psychological Bulletin, 101, 428–442. Hedges, L. V. (1981). Distribution theory for Glass’s estimator of effect size and related estimators. Journal of Educational Statistics, 6, 107–128. Hedges, L. V. (1982). Estimation of effect size from a series of independent experiments. Psychological Bulletin, 92, 490–499.

686

A. D. Daiuto et al.

Hedges, L. V., & Becker, B. J. (1986). Statistical methods in the meta-analysis of research on gender differences. In J. S. Hyde & M. C. Linn (Eds.), The psychology of gender: Advances through meta-analyses (pp. 14– 50). Baltimore: Johns Hopkins University Press. Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis . Orlando, FL: Academic Press. *Himadi, W. G., Cerny, J. A., Barlow, D. H., Cohen, S., & O’Brien, G. T. (1986). The relationship of marital adjustment to agoraphobic treatment outcome. Behavior Research and Therapy, 24, 321–332. Holmes, J. (1982). Phobia and counterphobia: Family aspects of agoraphobia. Journal of Family Therapy, 4, 133–152. Jacobson, N. S., Follette, W. C., & Elwood, R. W. (1984). Outcome research on behavioral marital therapy: A methodological and conceptual reappraisal. In K. Hahlweg & N. S. Jacobson (Eds.), Marital interaction: Analysis and modification (pp. 113–129). New York: Guilford. *Jannoun, L., Munby, M., Catalan, J., & Gelder, M. (1980). A home-based program for agoraphobia: Replication and controlled evaluation. Behavior Therapy, 11, 294–305. King, L., & Emmons, R. (1990). Conflict over emotional expressiveness: Psychological and physical correlates. Journal of Personality and Social Psychology, 56, 864–877. Kitch, D. P. (1983). Agoraphobic severity, dyadic adjustment, and marital structure: Patterns of change during treatment . Unpublished doctoral dissertation, California School of Professional Psychology, Fresno. Kleiner, L., & Marshall, W. L. (1987). The role of interpersonal problems in the development of agoraphobia with panic attacks. Journal of Anxiety Disorders, 1, 313–323. *Kleiner, L., Marshall, W. L., & Spevack, M. (1987). Training in problem-solving and exposure treatment for agoraphobics with panic attacks. Journal of Anxiety Disorders, 1, 219–238. Lange, A., & van Dyck, R. (1992). The function of agoraphobia in the marital relationship. Acta Psychiatrica Scandinavica, 85, 89–93. Last, C. G., Barlow, D. H., & O’Brien, G. T. (1984). Precipitants of agoraphobia: Role of stressful life events. Psychological Reports, 54, 567–570. Locke, H. J., & Wallace, K. M. (1959). Short marital adjustment and prediction tests: Their reliability and validity. Marriage and Family Living, 21, 251–255. *Mannetter, R. S. (1989). Comparison of two therapy modalities in the treatment of female agoraphobics . Unpublished doctoral dissertation, University of Iowa, Iowa City. Markman, H. J. (1991). Constructive marital conflict is NOT an oxymoron [Special issue: Negative communication in marital interaction: A misnomer?]. Behavioral Assessment, 13, 83–96. Mehta, M. (1990). A comparitive study of family-based and patient-based behavioral management in obsessive-compulsive disorder. British Journal of Psychiatry, 157, 133–135. *Milton, F., & Hafner, J. (1979). The outcome of behavior therapy for agoraphobia in relation to marital adjustment. Archives of General Psychiatry, 36, 807–811. *Monteiro, W., Marks, I. M., & Ramm, E. (1985). Marital adjustment and treatment outcome in agoraphobia. British Journal of Psychiatry, 146, 383–390. Norton, R. (1983). Measuring marital quality: A critical look at the dependent variable. Journal of Marriage and the Family, 45, 141–151. Ogles, B. M., Lambert, M. J., Weight, D. G., & Payne, I. R. (1990). Agoraphobia outcome measurement: A review and meta-analysis. Psychological Assessment: A Journal of Consulting and Clinical Psychology, 2, 317– 325. O’Leary, K. D. (1987). Assessment of marital discord . Hillsdale, NJ: Erlbaum. O’Leary, K. D., Riso, L. P., & Beach, S. R. H. (1990). Attributions about the marital discord/depression link and therapy outcome. Behavior Therapy, 21, 413–422. Olson, D. H., Sprenkle, D. H., & Russell, C. S. (1979). Circumplex model of marital and family systems: I. Cohesion and adaptability dimensions, family types, and clinical applications. Family Process, 18, 3–28. *Peter, H., & Hand, I. (1988). Patterns of patient-spouse interaction in agoraphobics: Assessment by Camberwell Family Interview (CFI) and impact on outcome of self-exposure treratment. In I. Hand & H.-U. Wittchen (Eds.), Panic and phobias 2: Treatments and variables affecting course and outcome (pp. 240– 251). Berlin: Springer-Verlag. Prager, K. J. (1995). The psychology of intimacy . New York: Guilford. Quadrio, C. (1984). Families of agoraphobic women. Australian and New Zealand Journal of Psychiatry, 18, 164–170. Rosenthal, R. (1984). Meta-analysis procedures for social research . Beverly Hills, CA: Sage. Ryle, A. (1966). A marital patterns test for use in psychiatric research. British Journal of Psychiatry, 112, 285– 293. Sabatelli, R. M. (1988). Measurement issues in marital research: A review and critique of contemporary survey instruments. Journal of Marriage and the Family, 50, 891–915.

BCT and Agoraphobia

687

Schroeder, B., Frank, M., Funfstuck, G., Schneider, S., & Hahlweg, K. (1996, November). Exposure in vivo in anxiety disorders: The impact of marital quality on treatment drop-out and treatment effectiveness . Paper presented at the annual meeting of the Association for Advancement of Behavior Therapy, New York. Shean, G. (1990). Interpersonal aspects of agoraphobia: Therapeutic implications. Psychotherapy in Private Practice, 8, 101–122. Sher, T. G., & Weiss, R. L. (1991). Negativity in marital communication: Where’s the beef? [Special issue: Negative communication in marital interaction: A misnomer?]. Behavioral Assessment, 13, 1–5. Shoham-Salomon, V., & Rosenthal, R. (1987). Paradoxical interventions: A meta-analysis. Journal of Consulting and Clinical Psychology, 55, 22–28. Smith, M. L., Glass, G. V., & Miller, T. I. (1980). The benefits of psychotherapy. Baltimore: Johns Hopkins University Press. Snyder, D. K. (1979). Multidimensional assessment of marital satisfaction. Journal of Marriage and the Family, 41, 813–823. Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and the Family, 38, 15–28. *Thomas-Peter, B. A., Jones, R. B., Sinnott, A., & Fordham, A. S. (1983). Prediction of outcome in the treatment of agoraphobia. Behavioural Psychotherapy, 11, 320–328. Wittchen, H., & Essau, C. (1991). The epidemiology of panic attacks, panic disorder, and agoraphobia. In J. Walker, G. Norton, & C. Ross (Eds.), Panic disorder and agoraphobia: A comprehensive guide for the practitioner (pp. 103–149). Pacific Grove, CA: Brooks/Cole. *Woods, J. (1989). The effects of spouse-based cognitive-behavior therapy on the treatment of agoraphobia . Unpublished doctoral dissertation, Western Michigan University, Kalamazoo.