The effect of an attachment‐oriented couple intervention for breast cancer patients and partners in the early treatment phase: A randomised controlled trial

Abstract Objective Patients and partners both cope individually and as a dyad with challenges related to a breast cancer diagnosis. The objective of this study was to evaluate the effect of a psychological attachment‐oriented couple intervention for breast cancer patients and partners in the early treatment phase. Methods A randomised controlled trial including 198 recently diagnosed breast cancer patients and their partners. Couples were randomised to the Hand in Hand (HiH) intervention in addition to usual care or to usual care only. Self‐report assessments were conducted for both patients and partners at baseline, postintervention (5 months), and follow‐up (10 months), assessing cancer‐related distress, symptoms of anxiety and depression, and dyadic adjustment. Patients' cancer‐related distress was the primary outcome. Results Cancer‐related distress decreased over time in both patients and partners, but the intervention did not significantly affect this decrease at postintervention (P = .08) or follow‐up (P = .71). A significant positive effect was found on dyadic adjustment at follow‐up for both patients (P = .04) and partners (P = .02). Conclusions There was no significant effect of the HiH intervention cancer‐related distress. The results suggest that most couples can cope with cancer‐related distress in the context of usual care. However, the positive effect on dyadic adjustment implies that the HiH intervention benefitted both patients and partners. Future studies should investigate how to integrate a couple focus in usual cancer care to improve dyadic coping in the early treatment phase.


| Participants
Eligible patients were women newly diagnosed with primary BC, who were ≥18 years, cohabited with a male partner, had no previous cancer diagnoses, had received no neoadjuvant treatment, had no history of hospitalisation due to psychosis, were able to read and speak Danish, and were not referred to or consulting any of the trial psychologists.
Partners had to be ≥18 years and be able to read and speak Danish.

| Enrolment
Eligible patients were identified and informed about the project during their hospital admission in relation to primary surgery. Enrolment was conducted at 3 Danish breast surgery departments from October 2011 to December 2012 for centres 1* and 2, † and April 2012 to January 2013 for centre 3. ‡ Consenting patients received additional information about the project by phone. If they consented to participate in the study, their partners were asked for verbal consent. Couples were randomised if completed questionnaires and signed consent forms had been returned.
Randomisation was stratified on centres, and each centre was block randomised. All except the independent statistician were blinded to block sizes and allocation sequence. Participants were for obvious reasons not blinded. Due to geographical reasons, it was not possible to randomise the psychologists to centres.

| Control condition: usual care
Usual care at all 3 centres consisted of verbal and written information on normal psychological reactions in relation to a cancer diagnosis. It was distributed by the local clinical staff.

| Intervention: HiH in addition to usual care
The HiH intervention consisted of 4 to 8 couple sessions led by a clinical psychologist up to 5 months after primary surgery. Attendance of both the patient and partner was required. The HiH intervention aimed to enhance dyadic adjustment through dyadic coping within the couples (eg, mutual understanding of attachment behaviour, perceived proximity and security, and creating new emotional experiences). The following issues should be addressed during couple sessions: couples' sense of attachment-related security, level of individual emotional distress and needs, knowledge of and experiences with cancer, psychological disorders, former stress-full life events, intimacy and sexual function, and other stressors. Enrolment implied 4 to 8 couple sessions, but the total number of couple sessions was decided by the couple and their allocated psychologist. All trial psychologists were experienced in working with therapeutic counselling of cancer patients and couples. Further details on the HiH intervention can be found in the published protocol article 20 and Appendix S1.

| Measurements
Data were obtained from a national database (time of primary surgery, T0) and self-assessment questionnaires at preintervention (T1), postintervention 5 months after surgery (T2), and follow-up 10 months after surgery (T3).

| Cancer-related distress
The Impact of Event Scale (IES) 21

| Therapeutic alliance
The "Bond" subscale from the Working Alliance Inventory-Short Revised assessed patients' and partners' perceptions of an affective bond between the psychologist and themselves. 24 These items were added to the questionnaire at T2 for participants in the intervention group. The scores range from 0 to 28.

| Additional support
At T2, all participants were asked if they had received any professional support and counselling (other than the intervention) from a doctor, nurse, psychologist, priest, social worker, or support group.

| Demographic and medical variables
Breast cancer characteristics were obtained from the Danish Breast Cancer Group-clinical database. 25 Cohabitation status and age were obtained from the Civil Registration System. 26  tion studies, 13 we included 199 couples.

| Statistical methods
Descriptive statistics were used to present demographic and diseaserelated variables at baseline. We used linear regression adjusted for baseline scores of the respective outcome and chemotherapy to test couples did not complete any sessions.
On average, patients in the intervention group perceived a powerful impact of BC in the intervention group assessed by IES (mean, 26.28).
Patients and partners in both groups scored relatively low on symptoms of anxiety and depression. Overall, dyadic adjustment increased in the intervention group, while it decreased in the control group.

| Primary outcome
We found a significant positive effect of the intervention on patients' cancer-related distress between the intervention and control group at T2 (P = .05), but after adjusting for baseline, the effect was nonsignificant (P = .08) with an effect size of −0.32 (Table 2). Adjusting for disease-specific variables had no significant effect on the primary outcome.

| Secondary outcomes
There was no significant effect on cancer-related distress for partners at T2 (P = .99) or for patients (P = .71) and partners (P = .27) at T3. Effect  These results substantiate that most patients and partners had a strong therapeutic alliance with the psychologist. However, there was not enough variance among respondents to perform mediation analyses. A total of 54 patients (56%) in the control group and 53 patients (52%) in the intervention group reported that they had received additional professional support (trial psychologists not included). This was 20 (26%) for partners in the control group and 33 (38%) for partners in the intervention group. This difference was mainly due to perceived support from nurses.

| DISCUSSION
This study did not confirm that a psychological attachment-oriented couple intervention could further decrease cancer-related distress than usual care. At T3, we found a significant effect of the intervention on dyadic adjustment for both patients and partners, while partners in the control group had a significant decrease in symptoms of depression.
The nonsignificant findings on cancer-related distress should be compared to similar studies. A German study including 72 couples coping with BC concluded that the significant differences in cancer-related distress were caused by baseline differences and not by a differential effect of the couple intervention. 29 These findings are in line with the results of our study that found a significant effect on distress at T2 for patients (P = .05), although not significant when adjusting for baseline values (P = .08). The steady decrease in cancer-related distress in the control group and the significant positive effect for partners on symptoms of depression at T3 in the control group suggest that the patients and partners can cope with both general and cancer-related distress in the context of usual care.
The significant effect on dyadic adjustment at T3 is in line with other couple intervention studies that found an effect on dyadic adjustment or marital satisfaction. 30,31 The German RCT of 72 couples found a larger albeit nonsignificant improvement in relationship satisfaction in the intervention group. 29 The divergent findings might partly be due to different conceptualisations of dyadic adjustment, eg, marital quality or relationship satisfaction.
The results on the primary outcome may be influenced by attrition (attrition rate: 26%-27% at T3), as indicated by an Australian couplebased study. 27 In our study, attrition could affect results into a more positive direction, if participants did not complete the questionnaires due to distress. The opposite might be the case if participants that did not complete the questionnaires did not feel burdened by the BC and found no reason to further participation. However, an analysis found no larger degree of cancer-related distress in dropouts compared to complete cases at baseline. Though, the significant effect on dyadic adjustment at T3 with no significant effect at T2 may be a chance finding due to attrition.
Our results could not confirm previous studies' recommendations that psychological couple interventions for cancer patients and partners should be offered during the early treatment phase. 32 We found that 92 of 250 couples (37%), whom had given consent to receive further information regarding the study, declined because they found it difficult to cope with a psychological intervention at that time point ( Figure 1). The timing may be more appropriate for partners, because both relationship challenges and challenges related to trying to offer support are present for them while patients are overwhelmed by disease-related concerns. 33,34 Study strengths include the randomised controlled design, the large sample size of 198 couples at baseline, the specified primary outcome, and the theoretical framework. Further, the intervention was developed specifically for our sample, the multicentre design increased a the generalisability, and the effects of the intervention were investigated in both patients and partners. Timing and content of the couple sessions were adapted for each couple within the limit of 4 to 8 sessions up to 5 months after primary surgery. Further, data on therapeutic alliance showed that the vast majority of couples perceived an alliance with the psychologist, which is an important prerequisite for a successful intervention. 35,36 We had access to detailed clinical information on each eligible case, which made it possible to adjust for baseline differences in patients' clinical situation.

| Study limitations
There is a risk of selection bias in the enrolment procedures followed at the centres. To ensure homogeneity in the screening of and information to eligible patients, clinical staff received written guidelines and coaching with the project manager. Further, attrition throughout the study may have influenced the results, and the initial participation rate of 26% may have decreased the generalisability of the findings.
There was no active control group to secure that any effects would be due to the attachment-oriented intervention and not merely due to attention from a psychologist. Finally, only 53 couples (52%) completed 4 to 8 couple sessions. However, empirical data obtained by trial psychologists suggested that expanding the period for couple sessions for more than 5 months would increase number of couple sessions.

| Clinical implications
This study adds important knowledge to the field of couple interventions in cancer. The results suggest that cancer patients and partners generally have a steady decrease in distress over time within the context of usual care. The effect on dyadic adjustment for both patients and partners should be investigated further, to enhance the focus on patients and partners as a dyad in clinical care. It would be interesting to investigate whether increased dyadic adjustment contributes to reduced cancer-related distress in the re-entry phase, being the phase, in which patients have to make the transition from treatment to early survivorship.