Study Design

Overall, 20 publications were included. Eight publications described eight RCTs,^{4,6,8,10,11,16–18} while 12 publications describe nine pre-post observational studies^{7,9,14,15,19–25} and one cohort study.²⁶ No clinical practice guidelines met the inclusion criteria.

Country of Origin

Three of the included RCTs were conducted in Canada,^4,11,17 two in Iran^8,16 and one in each of Germany,¹⁰ the United States¹⁸ and both Canada and the United States.⁶ Three of the pre-post observational studies were conducted in Canada,^22,23,25 two in the United States,^7,9,14,15 and one in each of Australia,¹⁹ the United Kingdom,²⁰ the Netherlands²¹ and both Germany and Austria.²⁴ The one included cohort study was conducted in the United States.²⁶

Patient Population

Each of the included studies included adults experiencing relationship distress, for varied reasons. Some studies included participants referring or self-referring for outpatient^{14–16,20,24} or inpatient²¹ couples therapy due to distress from any or an unspecified cause. Other studies included participants with distress from a specific cause, for example one partner suffers from obsessive compulsive disorder,^7,9 post-traumatic stress disorder,⁶ alcoholism,²⁶ borderline personality disorder,⁸ history of intrafamilial child abuse,¹¹ insecure attachment,^23,25 or an unresolved emotional injury.²² Others included couples where one partner is a pathological gambler,¹⁷ recently received a cancer diagnosis,^5,19 is facing end-stage cancer,⁴ or recently disclosed an affair.¹⁰ The average ages of participants varied between studies, with a low of 33.1 years in a pre-post study including couples with one member with obsessive compulsive disorder^7,9 to a high of 64 years in a pre-post study including couples referred for therapy following a cancer diagnosis.¹⁹ Most studies included participants with an average age in their 30’s or 40’s. Similarly, the average length of relationships among included participants varied from a low of seven years⁶ to a high of 27 years.⁵

Intervention and Comparators

The approach to couples therapy varied widely between included studies. Of the eight included RCTs, two assessed Emotionally Focused Couples Therapy,^4,11 and one each assessed Congruence Couples Therapy,¹⁷ Hope and Forgiveness Focused Therapy,¹⁶ Couple Dialectical Behavioural Therapy,⁸ Cognitive Behavioural Conjoint Therapy,⁶ Traditional Behavioural Couples Therapy,¹⁰ and Conjoint Intimacy Enhancing Therapy.⁵ Of the nine pre-post observational studies, two assessed Emotionally Focused Couples Therapy,^23,25 and one each assessed Cognitive Behavioural Conjoint Therapy,^7,9 Cognitive Existential Couples Therapy,¹⁹ Psychodynamic Psychotherapy,²⁰ psychotherapy,²¹ a combination of behavioural therapy or problem-focused therapy^14,15 and a variety of forms of couples therapy within the same study.²⁴ The one cohort study assessed Behavioural Couples Therapy.²⁶ Appendix 4 includes brief description of each of the included therapy types.

Among the RCTs included in the review, five used a waiting list control ranging from 12 weeks^6,10,17 to 24 weeks in duration.¹¹ The length of the waiting list was not reported (NR) for one study.⁸ One RCT used a control group in which the participants received no intervention¹⁶ and another used a control group in which the participants received usual care through a psychosocial oncology and palliative care program.⁴ Among the pre-post observational studies, participants served as their own control. The one cohort study in which the participants were couples with a female alcoholic included a demographically matched control group selected from a population of non-alcoholic males and females.²⁶

Outcomes Measured

Many outcomes were measured in the included studies, and summarized within this review, all related to relationship distress and psychosocial health.

Seven of the eight included RCTs measured some aspect of relationship satisfaction using either the Dyadic Adjustment Scale,^4–6,11,17 Perceived Relationship Quality Components Inventory,⁸ or the Partnership Questionnaire.¹⁰ Other relationship centred outcomes measured within the RCTs include:

• marital partner function (Systemic therapy Inventory of Change)¹⁷
• caregiver burden (Relationship-focused Coping Scale)⁴
• relationship intimacy (Personal Assessment of Intimacy in Relationships)⁵
• communication (Communications Pattern Questionnaire)⁵
• infidelity (unvalidated instrument)¹⁰
• self- and partner disclosure of thoughts, information and feelings (unvalidated instrument)⁵ and
• responsiveness (unvalidated instrument).⁵

Psychosocial outcomes measured within RCTs included:

• mental stress (Brief Symptom Inventory)¹⁷
• interpersonal cognitive distortions (Interpersonal Cognitive Distortions Scale)¹⁶
• trauma symptoms (Trauma Symptom Inventory)¹¹
• depression (Beck Depression Inventory)^4,6,10
• hopelessness (Beck Hopelessness Scale)⁴
• anxiety (State Trait Anxiety Inventory)⁶
• event-related distress (Impact of Event Scale)^5,10 and
• psychological distress (Mental Health Inventory).⁵

Six of the pre-post studies measured relationship satisfaction using either the Dyadic Adjustment Scale,^7,9,22,23,25 Martial Satisfaction Inventory²⁴ or the Quality of Marriage Index.^14,15 Other aspects of relationship quality and function that were assessed include:

• attachment anxiety (Experiences in Close Relationships, Attachment Injury Measure)^23,25
• trust (Relationship Trust Scale, Trust Scale)^5,22,25
• communication (Communication Patterns Questionnaire)^7,9
• marital function (Family Relationship Index, Golombok Rust Inventory of Marital State)^19,20
• relatedness (Personal Relatedness Profile)²⁰
• problem-solving (Interactional Problem Solving Questionnaire)²¹
• forgiveness (Enright Forgiveness Inventory, Forgiveness Measure, Interpersonal Relationship Resolution Scale)^22,25
• resolution of unfinished business (Unfinished Business Empathy and Acceptance, and Feelings and Needs, scales)²² and
• couple specific problems (Target Complaints Discomfort and Change Scale).²²

Psychosocial outcomes that were measured within the eight pre-post observational studies include:

• depression (Beck Depression Inventory, Hamilton Rating Scale for Depression, Center for Epidemiological Studies Depression Scale)^7,9,24
• psychological distress (Mental Health Inventory, Symptom Check List, Global Symptom Index)^19,21,22
• emotional control (Courtlauld Emotional Control Scale)²³
• event-related distress (Impact of Event Scale)¹⁹
• coping (Brief Cope, Benefit Finding Scale)¹⁹ and
• psychological state (Clinical Outcomes in Routine Evaluation).²⁰

The one included cohort study assessed aggression and violence (Conflict Tactics Scale) as an outcome relevant to this review.²⁶

Randomized Controlled Trials (RCTs)

The RCTs included in this review were of mixed quality. Some RCTs included design features such as blinding data analysts⁴ and adverse event tracking,^4,6 and their reports included explicit descriptions of study methods to allow a comprehensive study appraisal. Others were poorly reported, and perhaps poorly conducted, for example omitting descriptions of important study processes such as randomization,^{5,6,8,10,11,16,17} allocation concealment,^{5,6,8,10,11,16,17} and intervention compliance.^8,16 Most of the included RCTs provided an explicit description of study objectives, hypotheses, eligibility criteria, outcomes and interventions, which allows for a thorough appraisal study quality relating to these elements. With most studies omitting details of randomization and allocation concealment, an assessment of the success of the randomization process to reduce or remove the influence of known and unknown confounders cannot be made. Similarly, only two of the RCTs included a power calculation and justification for the sample size,^4,6 as evidence for the ability of the study to detect a clinically meaningful difference in measured outcomes. These two RCTs likewise included an analysis of clinical significance for primary outcomes, as did one further RCT that did not include a sample size calculation within the report.¹⁰ For the remaining five RCTs that did not include a power calculation, nor an analysis of clinical significance, it remains unclear as to whether these studies were sufficiently powered to detect meaningful differences. Of note, two of these five studies were small pilot studies that were not intended for hypothesis testing.^5,17 Two RCT reports included an explicit description of adverse events and a monitoring process,^4,6 while the remaining RCT reports do not discuss safety. In these cases it is unclear whether adverse events were not monitored, not reported, or did not occur.

All but one report provided an explicit description of the sampling process,⁸ which raises the potential for selection bias within this RCT. All but two reports included a description of the therapist(s) and training related to providing the study intervention,^8,16 raising the potential for variation within intervention delivery within these studies and therefore the ability to associate the intervention with observed outcomes. Intervention compliance was adequate in four of the RCTs,^4,6,11,17 not reported in two,^8,16 and low in two.^5,10 Low intervention compliance reduces internal validity of study results, as the ability to measure relevant outcomes associated with the intervention decreases. An intent to treat analysis was completed in three RCTs^5,6,10. For the remaining studies, non-compliant patients were excluded from the analysis raising the potential for overestimating treatment results.

While validated outcome questionnaires were used to assess outcomes across all included RCTs, the potential for social desirability bias cannot be ruled out with these self-report measures. Further, due to the nature of the intervention, blinding of patients or therapists is not possible; although in one study personnel who entered and checked data were blind to group assignment.⁴ The lack of blinding of patients and therapists across all included studies, in addition to the use of self-report measures, increases the potential for bias in outcome assessment, in particular since participants are aware that they received the intervention and the desired direction of effect.

Observational Studies

As with the RCTs, most of the included observational studies were of mixed quality. Two common study design features within this group are the non-randomized, uncontrolled nature of the nine included pre-post studies,^{7,9,14,15,19–25} and the non-randomized nature of the one included cohort study.²⁶ The lack of randomization increases the potential for selection bias, while the uncontrolled nature of the pre-post studies makes it impossible to distinguish intervention effects from other effects such as regression to the mean, natural progression, or social desirability. As with the RCTs, validated outcome measures were used across all observational studies, but due to the subjective nature of these self-report instruments, especially within a non-blinded, non-randomized, uncontrolled design, the potential for measurement error is increased. Finally, due to the nature of the intervention, neither blinding of patients nor therapists was possible. The lack of blinding, in addition to the use of self-report measures, increases the potential for bias in outcome assessment, in particular since participants are aware they received the intervention and the desired direction of effect.

As with the RCTs, most observational studies included an explicit description of study objectives, hypotheses, eligibility criteria, outcomes, and interventions, allowing for a comprehensive appraisal of study quality. Further, an explicit description of the therapist(s) and their related training across all studies provides assurance the intervention was delivered consistently and as intended.

None of the included studies, however, provided a power calculation nor justification for the number of included couples, although three studies were identified as pilot studies,^7,9,19,20 where hypothesis testing was not the main goal. Despite not providing justification for the sample size, four of the included observational studies included an analysis of clinical significance of the primary outcome, suggesting these studies were adequately powered to detect a meaningful difference.^{14,15,20,23,24} As with the RCT reports, none of the observational studies included information about adverse event tracking raising the potential that these important outcomes were not tracked, as opposed to not reported. External validity is further limited in five of the pre-post studies due to poor reporting of sampling procedures, in particular whether people who agreed to participate were different in any meaningful way from those who did not participate.^{7,9,19–21,24}

Intervention compliance was adequate within seven of the ten included observational studies,^{7,9,19,20,22,23,25,26} but low within two.^14,15,24 Compliance was not reported within one study report.²¹ For each of the observational studies, non-compliant patients were excluded from the analysis, which raises the likelihood of overestimating treatment results especially for those studies with low compliance.^14,15,24 In one pre-post study compliance was low within both study sites, but compliance rates differed significantly between study sites.^14,15 In addition, for this particular study, the publication notes considerable differences in participant characteristics across study sites in terms of ethnicity, religion, education and income, in addition to differences in intervention delivery in terms of focus, scope, and treatment duration.^14,15 Given results for this study were combined across study sites, these between site differences increase the potential for bias in outcome measurement since measurement effects will be impacted differently by both compliance and the intervention across included participants.

Relationship Satisfaction

Results were mixed among the seven RCTs that measured some aspect of relationship satisfaction.^{4–6,8,10,11,17} In two studies, a statistically significant improvement was reported between pre- and post-treatment between the intervention and wait list control groups for pathological gamblers (P < 0.01),¹⁷ couples self-referring for treatment due to a history of intrafamilial child abuse in the female (P < 0.05 for both couples and females only)¹¹ and for PTSD patients (change in Dyadic Adjustment Scale 12.22, 95% confidence interval [CI]: 5.72 to 18.72, P = NR) but not their partners (change in Dyadic Adjustment Scale 3.23, 95% CI: −2.35 to 8.81, P = NR).⁶ Two further RCTs reported statistically significant differences between treatment and control groups immediately post-treatment for people diagnosed with end-stage cancer (P < 0.0001)⁴ and 1 month post treatment among couples referred for therapy for borderline personality disorder in the male (P ≤ 0.01 for all but the trust subscale of the Perceived Relationship Quality scale).⁸ Three RCTs included an analysis of clinical significance and each reported a greater proportion of couples within the treatment group observing a clinically meaningful improvement as compared to couples in the control group.^4,6,11 Additionally, two studies measured the duration of effect by re-assessing relationship satisfaction 3-months post-treatment.^4,6 In both cases improvements in relationship satisfaction were maintained at the follow-up assessment.

Two RCTs reported non-statistically significant changes in relationship satisfaction following couples therapy as compared to a wait list¹⁰ or usual care⁵ control group. In one RCT including couples self-referring for therapy due to a recently disclosed affair,¹⁰ the difference in effect size pre-post intervention between the treatment group and a wait list control was not statistically significant for either the deceived (P = 0.225) or unfaithful (P = 0.141) partners. The other RCT included couples in therapy due to a recent diagnosis of prostate cancer and no statistically significant change was reported for either men diagnosed with prostate cancer or their partners.⁵

Relationship satisfaction was also measured in six of the pre-post studies where consistent improvements were documented.^{7,9,14,15,22–25} Differences in relationship satisfaction pre- and post- treatment were documented for couples in therapy due to insecure attachment (P < 0.001)²³ for OCD patients (P < 0.01) and their partners (P < 0.05),^7,9 for couples in therapy due to unresolved emotional injury for both the injured partner (P < 0.001) and the injurer (P < 0.001),²² and for both males (P < 0.001) and females (P < 0.001) self-referring for therapy.^14,15 In this last study, improvements were greater for those with greater levels of pre-counselling distress.^14,15 One study demonstrated mixed results with statistically significant improvements pre- and post-intervention for couples self-referring for therapy for some subscales of the Marital Satisfaction Inventory (P < 0.001 for each of global distress, affective communication, problem-solving communication, time together) but not others (P = not significant [NS] for each of sexual dissatisfaction, role orientation, family history of distress, dissatisfaction with children, conflict over child rearing).²⁴ Results were mixed in terms of the long term nature of improvements in relationship satisfaction. In the study including OCD patients and their partners, improvements were maintained 6 months post-treatment for OCD patients (P < 0.01) but not for their partners (P = NS). At 12 months, changes were nearly significant among OCD patients (P = 0.053) but again not significant for their partners.^7,9 Among couples with an unresolved emotional injury, post-treatment improvements were not maintained over the 3 month follow up (P = NR),²² but for a cohort of couples self-referring for therapy due to attachment injury, post-treatment improvements were maintained 3-years post-treatment (P < 0.005).²⁵

Other Relationship Outcomes

A range of other relationship-related outcomes were assessed across the seven RCTs included in this review, each assessed in one RCT:

• Marital partner function: Among couples with a pathological gambling member, a statistically significant difference in improvement in systemic function was observed between treatment and control groups post-intervention (P = 0.023) that was not maintained after 8-weeks of follow up (P = 0.054).¹⁷
• Caregiver burden: A statistically greater improvement in relationship-focused coping was observed among people in therapy due to a recent diagnosis of end-stage cancer as compared to a usual care control group (P = 0.02), but no difference was observed between groups in terms of demand (P = 0.88) or difficulty (P = 0.09) of caregiver burden.⁴
• Relationship intimacy, communication, self- and partner disclosure, responsiveness: One RCT including couples faced with a recent diagnosis of prostate cancer examined several aspects of relationship function, including intimacy, communication, self- and partner-disclosure and responsiveness. No significant differences were observed between the treatment and usual care control group on either of these measures.⁵

Among the observational studies, a similar range of relationship-related outcomes were assessed. For example, the one cohort study included in this review assessed aggression and violence using the Conflict Tactics Scale.²⁶ In this study, significantly less verbal aggression, overall violence and severe violence was observed within the intervention group that included couples in counselling for alcoholism in the female as compared to a matched control group. These significant differences applied for both male-to-female and female-to-male violence.

Trust was assessed in three pre-post studies.^22,23,25 In one study including couples self-referring for therapy for an unresolved emotional injury, an improvement in trust was observed immediately following treatment for the injured partner (P < 0.05) but not for the injuring partner (P = NS).²² The improvement for the injured partner was not maintained at a 3-month follow up.²² In a long term follow up study, assessment of trust at 3-years post-treatment suggested that couples who observed an improvement in their attachment injury during treatment had greater trust after 3 years as compared to those who did not resolve their attachment injury.²⁵ In this study injured partners reported a lower level of trust than did their offending partner at the 3-year follow up.²⁵ Finally, a study including couples in therapy due to insecure attachment reported that trust is not a significant predictor of relationship satisfaction, nor the rate of change of relationship satisfaction while in therapy.²³

Two pre-post studies assessed attachment anxiety.^23,25 One study including couples in therapy for insecure attachment observed that attachment anxiety is a significant predictor of improvement in relationship satisfaction, such that individuals with a higher level of attachment anxiety were more likely to observe an increase in marital satisfaction post-therapy.²³ In this same study, it was reported that attachment avoidance does not predict relationship satisfaction.²³ In the same long-term follow up study described above, improvements in attachment injury were maintained 3-years after treatment for those who had resolved their attachment injury and people with unresolved attachment injuries reported higher levels of avoidant attachment.²⁵

Marital function was assessed as an outcome in two pre-post studies and neither reported a significant improvement following therapy.^19,20

Two studies assessed forgiveness. In one study among couples self-referring for therapy for an unresolved emotional injury, forgiveness improved significantly for the injured partner only immediately post-treatment (P < 0.001), but the improvement was not maintained through the 3 month follow up period.²² No changes in forgiveness were observed among the injuring partner. In a long-term follow up study, injured partners who had resolved their attachment injuries with therapy reported more forgiveness than injured partners who did not resolve their injury at a 3-year follow up.²⁵ In this study, forgiveness was only assessed in the injured partner.

Other relationship outcomes were assessed in one pre-post study each:

• Communication patterns: In one study including couples self-referring to couples therapy due to a member with obsessive compulsive disorder (OCD), significant improvements were observed among OCD patients immediately post-treatment for both constructive communication (P < 0.05) and demand/withdrawal communication patterns (P < 0.05). For demand/withdrawal communication only, the improvements were maintained at both 6 months (P < 0.05) and 12 months (P < 0.01) post-treatment. No significant improvement was observed in terms of avoidance/withholding communication. For the partners of OCD patients, constructive communication was significantly improved immediately post-treatment as well as after 6 and 12 months post treatment (P < 0.001 for all). Avoidant communication patterns improved significantly following treatment (P < 0.05), and the improvements remained through 6 months post-treatment (P < 0.05) but after 12 months, there was no change from baseline (P = NS). Demand/withdrawal communication patterns did not significantly improve among partners with treatment.^7,9
• Relatedness: One study that employed a time-limited approach to therapy reported a significant improvement post-therapy on both scales of depressive and paranoid schizoid relatedness.²⁰
• Problem solving capacity: In one study including couples referred for inpatient couples therapy, problem solving capacity increased with treatment for people with a functional model of others and improvements were maintained at 6 and 18 months post-treatment (P = 0.01). For people with a dysfunctional model of others, problem solving capacity also increased during treatment but at a lower level, and improvements were not maintained at 6 and 18 months post-treatment (P = 0.14).²¹
• Resolution of feelings and needs: In the study including couples self-referring for therapy for an unresolved emotional injury, the resolution of feeling and needs improved for both the injured (P < 0.001) and the injuring partner (P < 0.05), as did feelings of acceptance and empathy (P < 0.001 for the injured partner; P < 0.05 for the injuring partner).²²
• Couple identified complaints: In the study described above, couples identified specific complaints at the beginning of the study to address in therapy. These were likewise improved post-treatment from the perspective of both the injured and injuring partner on scales that measure both discomfort (P < 0.001) and change (P < 0.001) related to the complaint.²²

Psychosocial Health Outcomes

Depression was the most common psychosocial outcome measured within the RCTs included in this review, being measured in three studies with mixed results.^4,6,10 In one study including couples faced with an end-stage cancer diagnosis, no difference between the intervention and the usual care control group was observed post-treatment (P = 0.46).⁴ Similarly, in a study including couples coping with PTSD no difference was found in depressive symptoms after treatment between the treatment and a waiting list control group.⁶ In another study including couples in counselling for a recently disclosed affair, improvements in depressive symptoms were observed in the deceived partners (P = 0.037) but not the unfaithful partners (P = 0.082).¹⁰

Depressive symptoms were assessed in two pre-post studies,^7,9,24 and in contrast to the RCTs consistent improvements were documented. In one pre-post study including couples self-referring for therapy significant improvements in depressive symptoms were observed immediately after therapy.²⁴ Similarly, in another pre-post study including couples with one partner with OCD, a significant improvement in depressive symptoms was observed pre- and post-intervention that was maintained through follow ups at 6 and 12 months (P for all comparisons < 0.001 as per the Beck Depression Inventory, and P = 0.02 for all comparisons as per the Hamilton Rating Scale for Depression).^7,9

Event-related distress was assessed in two RCTs^5,10 and one pre-post study.¹⁹ In one RCT including couples in counselling for a recently disclosed affair, significant improvements were observed in terms of the intrusion subscale of the Impact of Events scale for both the deceived and unfaithful partners (P = NR), and for the unfaithful partner only in terms of hyperarousal (P = NR).¹⁰ In a second RCT, no significant change was found between a treatment and usual care group in terms of event-related distress for couples faced with a recent cancer diagnosis.⁵ In one pre-post study including couples referred for therapy following a recent cancer diagnosis, a significant improvement was found between pre-and post-treatment for the avoidance (P = 0.021) and hyperarousal (P = 0.019) subscales of the Impact of Events Scale, but not the Intrusion subscale (P = NS).¹⁹

Psychological distress was assessed in three pre-post studies, with mixed results.^19,21,22 In one study including couples referred for inpatient couples therapy, among people with functional and dysfunctional models of others psychopathology improved with treatment.²¹ The increase was different between these groups at post-treatment only (P = 0.04) but not during treatment (P = 0.38). In another study including couples faced with a cancer diagnosis, no change in psychological distress was observed pre- and post-treatment.¹⁹ In a further study including couples in therapy for an unresolved emotional injury, significant improvement in terms of psychological distress was observed for the injured partners immediately post-treatment (P < 0.001), a change that was not maintained at a 3-month follow up. No change was observed among the injuring partners.²²

Other psychological outcomes were assessed in one RCT each:

• Hopelessness: In one study including couples faced with an end-stage cancer diagnosis, no difference between the intervention and the usual care control group was observed post-treatment in terms of feelings of hopelessness (P = 0.24).⁴
• Trauma-related symptoms: In a study including couples in therapy for the female with a history of intrafamilial child abuse, no difference was observed between a treatment and waiting list control groups in terms of trauma-related symptoms.¹¹
• In a study including couples with a pathological gambling partner, significant differences in improvement in mental distress were observed between people in a treatment and wait list control group immediately post-treatment (P = 0.001) and after an 8 week follow up (P = 0.035).¹⁷
• Interpersonal cognitive distortions: In one study including couples referred to pre-divorce counselling, interpersonal cognitive distortions were assessed for couples enrolled in a hope- and forgiveness-focused therapy group, a hope-focused therapy group, a forgiveness-focused therapy group, and a group who received no treatment. Improvements were observed among the combined therapy group as compared to the control group (P = 0.05), but not between the hope-focused group (P = 0.85) or the forgiveness focused group as compared to the control group (P = 0.74).¹⁶
• Anxiety: In a study including couples coping with PTSD, anxiety as measured by the State-Trait Anxiety Inventory decreased significantly among the intervention group (−10.60, 95% CI: −19.04 to −2.16, P = NR), but not the control group (0.84, 95% CI: −4.40 to 6.08, P = NR).⁶
• Anger: In the above study, anger expression decreased significantly among the treatment group (−8.02, 95% CI:−12.63 to −3.42) but not the wait list control group (−1.16, 95% CI: −4.55 to 2.23).⁶
• Psychological distress: In a study including couples faced with a recent cancer diagnosis, no significant difference in terms of psychological distress was found between couples in a treatment versus a usual care group.⁵

Finally, outcomes assessed in one pre-post study each included:

• Emotional control: In a study including couples referred to or self-referring for couples therapy due to insecure attachment, it was found that emotional control is not a significant predictor of relationship satisfaction, nor the rate of change of relationship satisfaction over the course of therapy.²³
• Coping: In the same study described above, no significant difference was found in terms of coping (P = NR).¹⁹
• Psychological function: One study that employed a time-limited approach to therapy measured overall psychological function and reported a non-significant decrease in the proportion of both males (P = 0.066) and females (P = 0.103) in terms of psychological function.²⁰