THE DEVELOPMENT AND EVALUATION OF AN EMPIRICALLY INFORMED INTERVENTION FOR PARTNERS OF WOMEN HOSPITALIZED FOR PERINATAL
DISTRESS
Matthew J. Cohen
A dissertation defense submitted to the faculty at the University of North Carolina at Chapel Hill in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the
Department of Psychology and Neuroscience
Chapel Hill 2018
ABSTRACT
Matthew J. Cohen: The Development and Evaluation of an Empirically Informed Intervention for Partners of Women Hospitalized for Perinatal Distress
(Under the direction of Don Baucom)
Perinatal distress (PD) is a term used to describe the mood, anxiety, or in rarer cases, psychotic symptoms that women experience during pregnancy or in the postpartum period. PD is associated with significant emotional and social impairments that impact women, their children, and their partners. When PD is more acute and safety concerns emerge for mothers and possibly their children, inpatient hospitalization may be indicated. Although hospitalization tends to result in improved immediate safety outcomes, many women are discharged when they are “better but not well.” Additionally, research shows that interpersonal and environmental factors, particularly those that exist within intimate relationships, can significantly help or hinder patient progress. Yet, the partner is also affected by the patient’s symptoms and often lacks the skills or
knowledge to support her in her recovery. Given the research demonstrating that integrating partners into treatment improves individual and relational outcomes across psychopathology, the aim of the present investigation is to address perinatal mood disorders through a dyadic lens by integrating partners into treatment. In this feasibility study, 20 male partners of women
significant gains in the context of their support self-efficacy, and patients reported significantly improved depression and anxiety symptoms. Patients also reported that they were significantly more satisfied with the support that their partners were providing, although the frequency of partner support behaviors was unchanged. The findings from this study suggest that the
TABLE OF CONTENTS
LIST OF TABLES………. vii
CHAPER 1: INTRODUCTION………...1
Causes of Depression, Anxiety, and Psychosis in the Perinatal Period………... 3
Consequences of an Untreated PD Episode………... 6
Impact of PD on Mothers ………... 7
Consequences of Maternal Antenatal and Postpartum PD on Children……… 8
Influence of Maternal PD on Intimate Relationships……… 9
Severe PD and Hospitalization……….. 10
Partner Involvement in Treatment………... 17
Partner Reactions to Psychopathology………. 18
The Partner Experience During Hospitalization……….. 26
Barriers to Couple Based Treatment Among Women with Severe PD………... 27
Relationship Education Initiatives…….……… 28
Core Elements of RE Interventions………. 30
Relationship Education for Individual Partners of Women Hospitalized for PD……….. 31
An Intervention for Partners During the Perinatal Period………... 32
CHAPER 2: CURRENT INVESTIGATION...………... 34
Study Hypotheses……….………..37
Cultivating and Integrating Stakeholders……….………..39
Participants.……… 39
Measures……… 39
Procedure………... 44
Introduction and Psychoeducation………... 45
Improving Partner Support Behaviors and Coping Strategies………. 46
CHAPER 4: RESULTS………....………... 50
Evaluate Sample Characteristics and Recruitment Capability.………..50
Clarifying Data Collection Procedures and Outcome Measures.………... 52
Examining Acceptability of the Intervention.………..………... ….. 53
Psychoeducation……….. 54
Communication Skills……….. 54
Self-Care and General Support……….... 55
Suggestions for Refining the Intervention………... 55
Evaluating the Context for Implementing the Intervention……….……… …..56
Preliminary Evaluation of Potential Intervention Effects………... 58
Partners’ Pre-Intervention to Four Week Follow Up Change………. 62
Within-Group Effect Size Comparisons……….. 65
CHAPER 5: DISCUSSION………....………... 68
Feasibility and Acceptability………...………... 68
Clinical Outcomes………...………... 72
Limitations and Future Directions………...………...…………... 78
LIST OF TABLES
Table 1 – Questionnaires administered across three time points………. 103
Table 2 – Client Satisfaction Questionnaire (CSQ-8)………...…..104
Table 3 – Individual characteristics of participating patients and partners………. 105
Table 4 – Characteristics of ineligible patients and reasons for not participating….………... 106
Table 5 – Patient pre-intervention scores to four-week follow-up scores………... 107
Table 6– Partner pre-intervention scores to four-week follow-up scores ……….. 108
Table 7 – Pre-intervention correlations for patients……… 109
Table 8 – Pre-intervention correlations for partners……… 110
Table 9 – Cross-partner baseline correlations……….. 111
Table 10 – Pre-intervention to four-week follow-up: depression, anxiety, & relationship satisfaction………... 112
Table 11 - Clinically significant and reliable change for patients………... 113
Table 12 - Clinically significant and reliable change for partners………...114
Table 13 - Effect sizes from pre-intervention to four-week follow-up………...….115
Table 14 - Recruitment rate data for interventions targeting perinatal distress……….…..116
CHAPTER 1: INTRODUCTION:
Societal expectations suggest that the perinatal period should be a time of unequivocal joy; yet, most women do not experience this time in such an absolute way. For many, the demands experienced in the perinatal period, coupled with other environmental and biological factors, can result in heightened psychological distress. Perinatal distress (PD) is a broad term used to describe a range of psychological symptoms experienced during pregnancy and in the year that follows childbirth (O'Hara & Swain, 1996). There is a diversity of presentations associated with PD, a disorder that can be characterized by high levels of anxiety (Staneva, Bogossian, Pritchard, & Wittkowski, 2015) or acute symptoms of depression (Rallis, Skouteris, McCabe, & Milgrom, 2014). In rarer cases, women with PD may experience psychotic
symptoms, characterized by delusions, hallucinations, cognitive disorganization, and general impairment (Spinelli, 2009).
Among the symptoms that women with PD experience, anxiety and depression are the most pervasive. Specifically, estimates suggest that 20% of women will exhibit depressive symptoms during pregnancy (Marcus, Flynn, Blow, & Barry, 2003), and approximately 19% of women will experience major or minor depression in their first three months postpartum (Gavin et al., 2005). A major depressive episode during the perinatal period mirrors an episode
Depressive symptoms have been shown to co-occur with anxiety across populations (e.g., Cully et al., 2006; Graff, Walker, & Bernstein, 2009), and this holds true for women in the perinatal period (Moss, Skouteris, Wertheim, Paxton, & Milgrom, 2009; Skouteris, Wertheim, Rallis, Milgrom, & Paxton, 2009). Research suggests that 25-45% of women will experience moderate to severe anxiety during the perinatal period, many of whom concurrently experience depression (Britton, 2008; Faisal-Cury & Rossi Menezes, 2007). When anxiety is the primary characteristic of PD, the collection of symptoms most commonly approximates the following anxiety disorders: Generalized Anxiety Disorder (GAD), Panic Disorder, and Obsessive
Compulsive Disorder (OCD). GAD, characterized by pervasive worry, occurs in 4-8% of women in the perinatal period. Panic Disorder, which is marked by recurrent panic attacks and persistent fear of experiencing future symptoms of anxiety, has a prevalence rate of 1-5% in the perinatal period (Wenzel, Haugen, Jackson, & Brendle, 2005). Finally, OCD is characterized by recurrent obsessional, unwelcome thoughts and resulting compulsive behaviors and occurs in
approximately 3-4% of women in the perinatal period. When OCD symptoms are experienced by mothers in the postpartum period, the obsessive thoughts often involve harming their children, contamination, or religious preoccupations (Jennings, Ross, Popper, & Elmore, 1999); common compulsions typically include avoidance, cleaning, checking, or other mental rituals (Sit,
Rothschild, & Wisner, 2006). Of note, although OCD and postpartum psychosis both involve the presence of bizarre thoughts, they are distinct in that OCD-specific intrusive thoughts are ego-dystonic; the mother experiencing the thoughts actively resists them, thereby causing high levels of distress and, subsequently, more intrusive thoughts (Abramowitz, Schwartz, Moore, &
As researchers and clinicians seek safe and effective treatment options for perinatal populations, it is imperative to understand the critical role that partners play in women’s recovery. The couples and psychopathology literature suggests that interpersonal processes are central in the development and maintenance of a range of psychological disorders (Fischer, Baucom, & Cohen, 2016), and this proposal integrates findings from research on couple functioning during the perinatal period in order to highlight the importance of including the partner in treatment for maternal PD. However, women who need to be hospitalized as a result of experiencing more severe PD may experience impairments that limit their ability to engage in couple therapy. As such, this proposal will focus on the development and implementation of a partner-focused intervention aimed at improving partner understanding and positive support behaviors.
Causes of Depression, Anxiety, and Psychosis in the Perinatal Period
In their efforts to understand how and why psychopathology develops, psychologists have historically drawn from the diathesis-stress model, which suggests that individuals who possess predisposing biological risk factors for a certain disorder are more likely to develop that disorder when exposed to environmental stressors (Robins & Block, 1989). When applied to women in the perinatal period, the diathesis-stress model suggests that biological vulnerability factors, coupled with stressful life events, can result in the emergence of PD symptoms (O'Hara, Schlechte, Lewis, & Varner, 1991). For example, in one study, women were assessed during pregnancy for three variables: depressive symptoms, whether they had an abuse history, and the presence of a functional polymorphism in their serotonin transporter gene. The findings indicated that women who had been subjected to abuse and possessed this polymorphism were at
sample who had either the gene or an abuse history only (Scheid et al., 2007). Similarly,
examinations of biological risk factors during pregnancy (e.g., high progesterone concentrations) have demonstrated that such factors do not directly predict depressive symptoms, but rather, have an indirect effect, with psychosocial stressors (e.g., lack of social support) and anxiety levels serving as mediators (Ross, Sellers, Gilbert Evans, & Romach, 2004). These studies, among others, have played an integral role in clarifying the complex dynamic between biology and environment that drives mental illness in the perinatal period.
From a biological and physiological perspective, research supports the notion that these factors play a role in the development of PD. For example, after a child is born, there is a well-established maternal decline in key reproductive hormones (e.g., estrogen and progesterone), a drop that is associated with the development of postpartum mood and anxiety symptoms (Bloch et al., 2000). In addition, the altered cytokine profile of women after delivery has been shown to contribute to these symptoms (Groer & Davis, 2006). Research also suggests that a past history of psychiatric illness, as well as a family history of psychopathology, increases the risk of developing perinatal mental health problems, findings that highlight this biological influence (Meltzer-Brody, Boschloo, Jones, Sullivan, & Penninx, 2013; Steiner, 2002). These associations, however, may also be explained by environmental factors (e.g., unstable family environment), suggesting that PD symptoms are likely not an entirely biological phenomenon in the context of depression and anxiety symptoms.
& Marcus, 2003). These findings, coupled with other genetic studies that have identified certain genes that play a role in the etiology of PD (Jones et al., 2007), suggest that this genetic effect is particularly relevant for women with postpartum psychosis.
In addition to genetic and biological risk factors, environmental stressors have also been shown to influence the development and maintenance of maternal PD. Pregnancy and labor itself can represent stressful life events (Kettunen, Koistinen, & Hintikka, 2016; Swendsen & Mazure, 2000), and the likelihood of PD developing increases when additional stressors are present. For example, from a gynecological perspective, an adverse labor and delivery event has been shown to increase the risk of maternal PD (Kettunen et al., 2016). In addition, factors associated with social and socioeconomic status, such as substandard living conditions or financial hardship, have been identified as significant environmental predictors of PD (Brown & Moran, 1997; Segre, O’Hara, Arndt, & Stuart, 2007). Many other prominent environmental stressors occur in the postpartum period and involve caring for the infant. These childcare stressors tend to involve disruptions in the child’s feeding and sleep schedules, as well as worries about the child’s health (Whiffen, 1992). Mother-infant interaction and infant temperament have also been proven to be strong predictors of postpartum distress symptoms, even when controlling for pre-existing maternal depression levels (Cutrona & Troutman, 1986). In this way, the presence of current stressful life events, as well as those that have occurred in the past (e.g., trauma or abuse histories), have been shown to increase a woman’s vulnerability to developing PD (Kendall-Tackett, 2007; Records & Rice, 2009; Silverman & Loudon, 2010).
Another key environmental stressor in the perinatal period involves the potential
transition for parents as they work to manage new roles while also navigating significant environmental demands (Levy-Shiff, 1999). This highlights the call for change that exists in intimate relationships and the importance of couple flexibility and cohesion during this time. When there are deficits in these areas, many couples experience difficulties effectively
transitioning to parenthood. In fact, over two-thirds of couples report a reduction in relationship adjustment after having their first child (Shapiro, Gottman, & Carrere, 2000). The presence of relationship distress, while normative, represents a significant environmental stressor. For example, women who are dissatisfied in their intimate relationship are more likely to experience depression and anxiety in the perinatal period (Graff, Dyck, & Schallow, 1991; O'Hara, 1986). A lack of perceived social support during the perinatal period has similarly been proven to portend negative mental health outcomes (Barnes, 2006; C. L. Dennis & Chung, 2006), with low levels of partner support identified as a key risk factor for PD (Milgrom et al., 2008). In addition, other negative relational qualities (e.g., high levels of partner criticism) have been shown to place women at greater risk for developing PD symptoms (Fisher, Feekery, & Rowe, 2002).
Consequences of an Untreated PD Episode
Given that women in the perinatal period are carrying or caring for an infant, the presence of maternal mental illness during this time can be more consequential than a similar diagnosis at other times during the life span. This section details the psychological and physical sequelae of PD for mothers, their children, and their intimate relationships (i.e., the relationship with a spouse or partner).
Impact of PD on mothers. If untreated, the presence of PD can result in significant psychological and behavioral consequences for mothers. From a psychological standpoint, PD is associated with additional comorbid psychopathology, both during the antenatal and postpartum period, including substance use disorders (Zuckerman, Amaro, Bauchner, & Cabral, 1989) and eating disorders (Mazzeo et al., 2006). Sleep disorders are also often comorbid with perinatal mental health disorders, from women who experience predominantly anxiety to those with psychosis (Hanley & Mintzes, 2014; Ross, Murray, & Steiner, 2005). In fact, poor sleep quality among women in the postpartum predicts depression in women with a history of Major
likelihood of regularly attending obstetrical visits (Bonari et al., 2004).1 In more acute cases, women with PD are at heightened risk of attempting suicide (Lindahl, Pearson, & Colpe, 2005).
Consequences of maternal antenatal and postpartum PD on children. The presence of psychological distress in the antenatal period also appears to negatively impact the health of the developing child, both in utero and beyond the gestational period. For example, rates of in-utero fetal movement are elevated in women with antenatal depression, (Dieter et al., 2001), an outcome associated with low birth weight and preterm delivery, the foremost causes of infant mortality in the United States (Grote et al., 2010). As mentioned earlier, substance use during pregnancy is more common among women with PD, which can similarly result in low birth weight, in addition to a host of neurobehavioral abnormalities (e.g., impairment in autonomic regulation, increased startle response; Behnke et al., 2013). Low birth weight also can be a product of insufficient maternal weight gain during pregnancy, another consequence of PD described above (Paarlberg et al., 2005). After delivery, infants of mothers who experience antenatal depression are less responsive, cry more often, and experience higher rates of sleep disturbance (Field, 2011).2 Similarly, mothers who report high levels of stress during pregnancy have children who are less likely to meet normative developmental milestones, and more likely to engage in antisocial behavior, and exhibit greater hyperactivity (Bonari et al., 2004). Although there may be third variable explanations for these findings, broadly, these results suggest that maternal depression and anxiety in the antenatal period impacts children in diverse and longstanding ways.
1 These deficits in self-care behaviors have also been shown to impact the developing fetus; specific implications
will be discussed in the following paragraphs.
The experience of psychological distress in the postpartum period presents a different, but similarly consequential, set of risks to mothers and infants. Primarily, mothers with PD are at risk for negative breastfeeding outcomes (e.g., breastfeeding difficulties, dissatisfaction with infant feeding; Dennis & McQueen, 2009), a concern both for the child and for the mother, given that breastfeeding is associated with reduced rates of maternal depression and improved physical health in the infant (Bier, Oliver, Ferguson, & Vohr, 2002; Figueiredo, Canário, & Field, 2014). Depressed mothers also tend to be more self-focused (Salmela-Aro, Nurmi, Saisto, &
Halmesmäki, 2001), less engaged with their newborns (Lovejoy, Graczyk, O'Hare, & Neuman, 2000), and generally more hostile and unresponsive (Flykt, Kanninen, Sinkkonen, & Punamäki, 2010). As a result, rates of child neglect and abuse are also greater in mothers with postpartum distress (Cadzow, Armstrong, & Fraser, 1999). Women with PD attend fewer well-child exams within the first two years of their child’s life (Gaffney, Kitsantas, Brito, & Swamidoss, 2014) and visit the emergency room with their infants significantly more (Zajicek-Farber, 2009). These findings suggest that mothers with postpartum distress experience cognitive and emotional deficits that make it difficult to fully and consistently attend to their children. More broadly, considerable research on women in the perinatal period demonstrates that children whose mothers suffer from PD are at risk across developmental time points, from infancy through adolescence.
can serve as a chronic stressor on relationships, which in turn often results in elevated conflict and distress (Mamun et al., 2009). Partners of women with PD tend to experience similarly diminished relational outcomes (Davey, Dziurawiec, & O’Brien-Malone, 2006), as well as heightened levels of anxiety and depression themselves (Bielawska & Kossakowska, 2006; Roberts et al., 2006). Conversely, diminished relationship functioning in the perinatal period has been shown to be a robust predictor of postpartum depression (Boyce & Hickey, 2005). In fact, the quality of a women’s intimate relationships during the perinatal period has been shown to predict the intensity and duration of distress symptoms, as well as risk for relapse (Beck, 2001). For example, in the context of postpartum depression, women who report higher levels of relationship distress endorse more severe symptoms, which occur for a greater duration of time (Fisher et al., 2002). Although PD may often be conceptualized as a form of individual
psychopathology, it is a disorder with sequelae that extend well beyond the mother. The impact of PD on intimate relationship functioning will be discussed in further detail in future sections. Severe PD and Hospitalization
developing infant at risk (Lindahl, Pearson, & Colpe, 2005). Mothers with postpartum psychosis who are actively psychotic are more likely to experience impairments in their ability to care for themselves, as well as deficits in cognition and judgment that may put their children at risk (Sit et al., 2006). Across these different presentations, women are hospitalized when their providers determine that high levels of risks are present.
Across psychiatric patients, admission to an inpatient unit tends to result in improved short-term safety outcomes (Glick, Sharfstein, & Schwartz, 2011); however, hospitalization alone does not represent sufficient treatment for high-risk patients. In general, the primary goals of psychiatric hospitalization are to improve immediate patient safety outcomes, which can occur through several different avenues. On inpatient psychiatric units, patients do not have access to means to do harm and are regularly monitored, depending on their level of risk. This context, coupled with the pharmacological and physical restraints that can be used during hospitalization, make it less likely that individuals are able to act on suicidal or homicidal ideation while
hospitalized (Jacobs & Brewer, 2004). Although these methods may be effective for reducing the likelihood of suicide attempts during inpatient admission, there is a lack of empirical data to suggest that these techniques reduce the risk of self-harm or suicidal behaviors after patients leave the hospital. While psychiatric hospitalization can improve immediate safety outcomes, the act of admission alone does not necessarily result in long-term symptom remission.
(Cooper, Willy, Pont, & Ray, 2007) and in the postpartum period (Miller, 2002). The use of antidepressants during the perinatal period has been associated with improved outcomes among individuals experiencing anxiety (Misri, Reebye, Milis, & Shah, 2006) and depression (Horowitz & Goodman, 2005). For women experiencing mania or psychosis in the perinatal period, Lithium and other mood stabilizers are often prescribed, given that antidepressants can trigger or
exacerbate manic symptoms (Yonkers et al., 2009). In the broader psychiatric population, mood stabilizers significantly outperform placebos in terms of preventing manic episodes, although they have not been shown to improve depressive episodes to the same extent (Geddes, Burgess, Hawton, Jamison, & Goodwin, 2004). Research also supports the use of antipsychotics in treating postpartum psychosis, although given the low prevalence rates of this disorder, the majority of studies in this area have been case reports (Doucet et al., 2011). Finally,
benzodiazepines are also often prescribed to women with PD in order to provide immediate symptom relief, specifically in the context of anxiety (Vythilingum, 2009). Benzodiazepines offer short-term relief that can be appealing to patients and providers, though the risk for
dependence as well as neonatal withdrawal syndromes suggest that they do not represent a long-term solution for most women (Vythilingum, 2009). The current research on the use of
psychotropic medication for PD indicates that they tend to result in improved short-term safety outcomes during hospitalization.
dictate that within most clinical samples, participants should be randomly assigned to an experimental group who receives the treatment and a control group who does not receive the treatment. However, ethical guidelines have historically precluded researchers from taking these steps with psychotropic medications and perinatal populations given the potential negative impact to the developing neonate or to the breastfeeding infant. As such, most of the
recommendations for pharmacologic treatment are drawn from the evidence on medication for non-perinatal samples (Horowitz & Goodman, 2005) or on retrospective studies of women who chose to take medication during the perinatal period. These retrospective studies are limited in that without a control group, patient progress may be attributed to a host of other factors unrelated to medication (e.g., passage of time, improved infant sleep). Beyond that, given the recommendation by the American Psychiatric Association as well as the American College of Obstetricians and Gynecologists that women experiencing mild to moderate depression in the perinatal period should consider psychosocial approaches as the first line of treatment, women who opt to take psychotropic medication in the perinatal period are more likely to experience greater symptom severity than those who choose to discontinue their medication (Yonkers et al., 2009).
Although the full scope of negative effects that psychotropic medications have on women and their children have not been fully explored, emerging research suggests that they do
(Kieviet, Dolman, & Honig, 2013). However, most of these PNA symptoms tend to remit within two weeks postpartum (Moses-Kolko et al., 2005). Further, use of certain SSRIs in the first trimester is also associated with increased risk of certain birth defects (e.g. septal defects with sertraline), although these risks are still relatively small (Louik, Lin, Werler, Hernández-Díaz, & Mitchell, 2007). There are also fetal effects of benzodiazepine use (e.g. neonatal toxicity) and as a result, they are not recommended in women’s first or third trimesters (Vythilingum, 2009). Meanwhile, among women who are breastfeeding, psychotropic medication is secreted in the mother’s breast milk (Weissman et al., 2004). However, infant exposure to antidepressants in breastmilk is minimal (Berle et al., 2004) and is not associated with developmental delays (Weissman et al., 2004) nor any significant functional effects in infants (Epperson et al., 2001). Finally, use of atypical antipsychotics during pregnancy is associated with an increased risk of preterm birth as well as major congenital malformations in the infant (Terrana, Koren,
Pivovarov, Etwel, & Nulman, 2015). In the postpartum period, certain antipsychotics (e.g., quetiapine, olanzapine) appear to confer few, if any, side effects to the infant while other categories (e.g., clozapine, amisulpiride) show considerable concentration in maternal breast-milk and are, therefore, contraindicated (Pacchiarotti et al., 2016).
2006). In fact, the physiological and biochemical makeup of newborns tend to mirror that of their mothers, suggesting that mothers with elevated cortisol and reduced dopamine and serotonin have newborns with similar profiles (Field et al., 2006).
The physiological effects of PD, coupled with the behavioral sequelae (e.g., reductions in self-care behaviors) that are associated with negative maternal, infant, and interpersonal
outcomes, indicate that psychotropic medication is often indicated for women hospitalized for severe cases of PD. Further, despite the lack of empirical research and potential side effects associated with psychotropic medication use during this time, the high-risk nature of women with severe PD suggests that stabilization during hospitalization is a priority.
Although antidepressants, sedatives, and antipsychotic medications may demonstrate efficacy for patients during hospitalization, their long-term value for treating women
experiencing severe PD is unclear. Specifically, women taking psychotropic medications for PD often experience significant residual symptoms, as well as inflated relapse rates (England & Sim, 2009). Beyond that, while psychotropic medications may improve maternal psychological
medication should not represent a stand-alone treatment (C. L. Dennis & Ross, 2006; O'Hara & Swain, 1996).
Another issue relating to medication in the perinatal period is that compliance tends to be inconsistent. Broadly, medication compliance among individuals with depression ranges from 40-70% (Myers & Branthwaite, 1992). Similarly, there is inconsistent outpatient medication compliance among women in the perinatal period due to the perceptions about risks that these medications may pose to neonates (C. L. Dennis & Chung, 2006). Among the women who fill their prescription, the majority do not take their medication as prescribed or discontinue use before the recommended time (Boath, Bradley, & Henshaw, 2004) . Qualitative research on antidepressant use in the perinatal period suggests that women who do faithfully take medication during this time experience heightened levels of guilt and anxiety (Bonari, Bennett, Einarson, & Koren, 2004). Stigma plays a role in driving these emotions (Goodman, 2009); indeed, taking medication as a result of heightening depression or anxiety while pregnant or postpartum
represents a departure from societal expectations that the perinatal period should be the happiest time in one’s life. Taken together, medication compliance is inconsistent in the general
population, and this issue is further magnified among women in the perinatal period who must contend with concerns about their child, as well as secondary emotions stemming from societal stigma.
Finally, from a broader systems-level, the current structure of insurance plans and psychiatric hospitalization in the United States is such that patients (both perinatal and non-perinatal) are often discharged when stabilized, even if they still experience symptomology that impairs their ability to function (Frank & Glied, 2006). In other words, patients may be
describing psychiatric hospitalization in the era of deinstitutionalization (Klerman, 1977). Indeed, although many patients who are discharged from inpatient settings experience a reduction in risk from their initial admission date, they often still experience acute illness (Sharfstein, 2009). The first week after discharge is a time of particularly high-risk for suicide for individuals who were hospitalized, and, yet, only half of patients visit an outpatient provider within a week of their discharge (Hunt et al., 2009). It is therefore imperative for clinicians in inpatient settings to continue to prioritize helping women with severe perinatal mental illness to find short-term stability while also working to integrate non-pharmacological methods of facilitating long term recovery.
Partner Involvement in Treatment
While psychopharmacology often represents the first line of treatment for women hospitalized for PD, there are many non-pharmacological strategies that, unlike medications, do not evoke concerns about implications for the neonate. One such approach involves engaging family members in treatment. Across psychopathology, interventions aimed at improving family dynamics and communication patterns have strong empirical support. For example,
individual treatment, interpersonal processes have consistently been proven to be an influential factor in improving psychological outcomes.
As noted earlier, intimate relationship functioning shares a bidirectional association with psychological distress, such that presence of one can precipitate the other (Whisman & Baucom, 2012). Most research in this area has focused on individuals with depression, and research shows that patients reporting higher levels of relationship distress at the outset of treatment experience reduced psychological outcomes as compared to those with lower levels of relationship distress (Denton et al., 2010). Similarly, patients who report higher levels of relationship distress at the end of individual treatment for depression are significantly more likely to report heightened symptom severity at an 18-month follow-up (Whisman, 2001). These findings are consistent across disorders, including bipolar disorder (Miklowitz & Johnson, 2009) as well as a range of anxiety disorders (Dewey & Hunsley, 1990; Zinbarg, Lee, & Yoon, 2007). This research demonstrates that relationship distress represents a chronic stressor that can be a factor in producing and worsening individual psychopathology, as well as reducing the efficacy of individual psychological treatments. Of course, one’s intimate relationship can also be a source of strength and serve as an important protective factor; the benefits of supportive relationships will be explored further in later sections.
Partner reactions to psychopathology. One relational factor that can play a significant role in the development and maintenance of psychopathology involves the partner’s reaction to the patient’s psychological symptomology. Often, well-intentioned partners take steps to reduce their partner’s distress and in doing so, inadvertently reinforce maladaptive patterns that
in anxiety that accommodating behaviors may serve, they also portend heightened OCD severity and reduced treatment outcomes (Boeding et al., 2013). Similarly, a partner of someone with depression may engage in behavioral accommodation (e.g., agreeing to do all of the housework so that the patient can stay in bed) which serves to both reduce patient distress in the moment, while also precluding patients from building self-efficacy and perpetuating long-term avoidance.
Whether or not accommodation is present, partners of patients who experience
psychopathology tend to feel burdened and resentful (Snyder & Whisman, 2003), which in turn increases the likelihood of partner criticism and hostility toward the patient (Hooley, 2007). Criticism, hostility, and other negative relational qualities (e.g., emotional over-involvement) are grouped together under the umbrella of expressed emotion (EE), a construct that captures many psychiatric patients’ family environments (Butzlaff & Hooley, 1998). Intimate partners of
individuals with psychopathology who rate more highly in EE tend to be less accepting and more likely to invalidate the patients’ experiences by suggesting that their symptoms are controllable (Butzlaff & Hooley, 1998). Since the introduction of EE in the late 1950s (Brown, Carstairs, & Topping, 1958), it has proven to be an important explanatory variable that has been successfully applied to a range of psychopathology, from depression and anxiety to psychosis (Hooley, 2007).
Primarily, EE has been examined as a factor associated with patient relapse and negative treatment outcomes (Butzlaff & Hooley, 1998). The research on EE initially focused on patients with schizophrenia, and findings within this literature consistently demonstrated positive
within behavioral therapy (Chambless & Steketee, 1999). Similar results have been found across depressive disorders where individuals with depression living in high EE environments are at increased risk for relapse (Butzlaff & Hooley, 1998). This link between EE and relapse is consistent across other disorders as well, including eating disorders, personality disorders, and substance use disorders (Hooley, 2007). Notably, there is a particularly strong association between EE and relapse among patients who are hospitalized most often with chronic psychological disorders (Butzlaff & Hooley, 1998). This finding highlights the bidirectional association that exists between psychopathology and EE; severe, chronic mental illness can result in heightened criticism and hostility among family members, and in turn, such
environments can erode feelings of patient self-efficacy and self-esteem, which serves to worsen psychopathology (Whisman & Uebelacker, 2009).
In addition to predicting relapse and other negative psychological outcomes, high levels of EE have also been shown to negatively impact medication compliance across disorders. For example, one study examined recently hospitalized patients with bipolar disorder whose families were high in EE and determined that level of EE served as a key predictor of medication
compliance (Miklowitz, Goldstein, Nuechterlein, Snyder, & Mintz, 1988). Findings within the psychosis literature mirror these results. Several studies have examined recently hospitalized first-episode psychosis patients, and the results suggest that critical or hostile family attitudes predict inconsistent medication compliance, in spite of the fact that the majority of patients in the study showed significant symptom improvements while on medication during their
While critical partners may portend relapse or other negative treatment outcomes, regular contact with family members who rate lower on EE (e.g., those who exhibit emotional warmth and support), can be protective (Bhugra & McKenzie, 2003). Indeed, across psychopathology, patients who report greater intimacy and higher levels of support from their partners tend to experience a more rapid recovery and generally improved psychological outcomes (Misri, Kostaras, Fox, & Kostaras, 2000). This protective effect appears to be especially salient in women who are vulnerable to PD, as higher levels of relationship satisfaction and partner support have consistently been found to be protective against depressive outcomes, both during pregnancy and the postpartum period (Milgrom et al., 2008; Pilkington, Milne, Cairns, Lewis, & Whelan, 2015). Similarly, there is a demonstrated negative association in the postpartum period between relationship satisfaction and maternal anxiety (Gambrel & Piercy, 2015), although this finding has not been replicated to the same extent among women experiencing heightened anxiety during pregnancy (Pilkington et al., 2015). Taken together, the literature on interpersonal relationships suggests that a partner can play an important role in influencing the development of PD symptomology.
qualitatively different in some ways for women in same-sex relationships. For that reason, this project will focus exclusively on opposite-sex relationships and explore men’s capacity to offer support during the perinatal period.
Broadly, compared to women, many men tend to exhibit relationship skill deficits that make them less effective at understanding their partner’s experience, thinking about their lives together through a relational lens, and offering quality support (Rankin-Esquer, Baucom, Clayton, Tomcik, & Mullens, 1999). For example, both men and women report greater
relationship adjustment when they are able to engage in reciprocal self-disclosures of personal thoughts and feelings within their relationship (Sprecher & Hendrick, 2004). Yet, men tend to offer significantly fewer self-disclosures than their female partners (Brody & Hall, 2010), endorse less comfort with emotional closeness (Feeney, 1994), and are less likely to experience feelings of affiliation or social bonding (Brody & Hall, 2010). The difficulties that many men experience in forming connection tend to be heightened in the perinatal period, a time
characterized by greater relational negativity and conflict, as well as reduced self-disclosure, praise, time for physical intimacy, and distraction-free communication (Doss, Rhoades, Stanley, & Markman, 2009).
In order to offer support to a distressed partner, there first needs to be recognition of the other person’s experience, an awareness that men are less likely than women to possess.
accounting for other people’s experience as well (Rankin-Esquer et al., 1999). These results are consistent with Acitelli’s (1992) earlier work on relationship awareness, a term used to describe one person’s ability to attend to their relational patterns and themselves in the context of their relationship, which indicated that women possess significantly more relationship awareness. This collection of literature suggests that men are less likely than women to view their lives through a relational lens.
In general, when men exhibit lower levels of RSP, their partners tend to report feeling less intimacy and trust within the relationship while also engaging in more negative
communication (e.g., blame, invalidation; Sullivan & Baucom, 2005). It stands to reason that deficits in relational awareness during the transition to parenthood, an inherently dyadic time, can magnify these negative relational patterns. Both during pregnancy and the postpartum period, women endure significant physical and emotional demands, while also typically serving as the infant’s primary caregiver after the child’s birth, even when both partners are working (Nomaguchi & Milkie, 2003). Given that infants depend on their caregivers for food, shelter, and 24-hour care, the postpartum period calls for partners to show greater awareness for the mothers’ experience. During this time associated with significant environmental stressors, partners who have a difficult time engaging in RSP may be less likely to recognize women’s need for support.
are unlikely to offer positive support when their partners are experiencing negative affect (Pasch, Bradbury, & Davila, 1997). In one study examining perceptions of support during times of stress, while men and women did not differ on their levels of support received, women reported that men’s support was consistently accompanied by negative behaviors (e.g., criticism, arguments; Neff & Karney, 2005). Similarly, in another study of support behaviors, researchers found that the quality of support (i.e., the extent to which the support behaviors correspond to the
disclosures of stress) that men offer in situations involving negative affect were especially low (Bodenmann et al., 2015). It appears that these negatively arousing interpersonal interactions tend to prompt emotional withdrawal in men (Gottman & Notarius 1988), a response often referred to as stonewalling, which can serve to erode feelings of intimacy and connection within relationships (Gottman, 1994). The broader findings on interpersonal relationship functioning suggest that negative behaviors more strongly predict mental health outcomes than positive behaviors (Vinokur & Van Ryn, 1993); therefore, the presence of men’s negative behaviors while offering support may undermine their attempts to be helpful. Given that stress and negative affect are prevalent among women during pregnancy and the postpartum period (Gavin et al., 2005; Marcus et al., 2003), men are more likely to withdraw emotionally and as a result,
experience reduced connection. Taken together, emotional withdrawal in male partners, coupled with other displays of negative communication, can be central contributors in the development and maintenance of PD (Marks, Wieck, Checkley, & Kumar, 1996).
There are a number of hypotheses that have been proposed to explain the difficulty that many men experience in offering emotional support, especially when their partners are enduring emotional difficulties. One explanation highlights the gender differences that exist at a
experience greater activity in hypothalamic-pituitary-adrenal axis (HPA) when exposed to environmental stressors (Kudielka & Kirschbaum, 2005). The intuitive response that many men experience in response to this heightened arousal, as noted earlier, is to emotionally withdraw from the stressor (Gottman & Levenson, 1988), which in this case is their partner. Men’s propensity to react to stressful interpersonal situations by turning inwards and away from
emotional sensitivity tends to be diametrically opposed to that of women, who are more likely to exhibit reduced egocentricity and greater awareness for others’ experience, sometimes
characterized as a “tend and befriend” response (Tomova, von Dawans, Heinrichs, Silani, & Lamm, 2014).
In addition to these physiological contributors, many men in western cultures have not historically been socialized in a way that promotes strong interpersonal relationships. While research on couple functioning demonstrates that healthy relationships are contingent on openness (Sprecher & Regan, 2002), men are more likely to be taught to hide their private experiences (Brooks, 1998). Similarly, men are socialized to be stoic and self-reliant (Addis & Mahalik, 2003; Pollack & Levant, 1998), qualities that run counter to the collaborative,
emotionally expressive person who is more likely to have strong intimate relationships (Epstein & Baucom, 2002; Lavee & Ben-Ari, 2004). Starting at age four, boys are more likely to incur judgment and punishment for expressing emotion and are also less likely than girls to be held (Fivush, 1989). These early messages persist throughout the lifespan and perhaps unsurprisingly, traditionally masculine ideals are inversely associated with emotional intelligence, relationship satisfaction, communication quality, and intimacy (Parker, Taylor, & Bagby, 2001).
relationships, especially when their partner is in distress. Of course, the perinatal period represents a period of physical and emotional stress and is commonly characterized by chronic anxiety and distress, which highlights the need for partner understanding and support during this time. Given the demanding nature of the perinatal period, the absence of these behaviors tends to be even starker than it might otherwise be at other developmental stages within a relationship.
Barriers to couple-based treatment among women with severe PD. Given the
importance of intimate relationships during the perinatal period and the difficulty that many male partners experience in offering support, as well as in coping with their own struggles, it would appear to be an opportune time for a couple-based intervention. Indeed, couple-based
Amaresha, Jangam, Muralidharan, & Jain, 2015). Taken together, findings suggest that women hospitalized for PD might benefit from greater cognitive and emotional stability before they are able to effectively engage in couple therapy.
Relationship Education Initiatives
Although couple therapy may not be as effective during psychiatric hospitalization for women with PD, partner understanding and support are crucial factors in the patient’s immediate recovery and long-term psychological well-being. Of note, research shows that interventions focusing on family members successfully reduce EE and cultivate greater self-esteem and coping skills within family members, which in turn improves patient outcomes (Kuipers et al., 2006). Given that, one possible avenue for intervention among women hospitalized for heightened PD is through relationship education (RE), an intervention designed to educate individuals and couples on ways to create and maintain healthy, mutually satisfying relationships (Hawkins, Blanchard, Baldwin, & Fawcett, 2008).
There are a range of RE initiatives that have undergone empirical investigation, both when participants attend individually and as a couple. Early RE interventions were exclusively dyadic and typically offered to committed couples in healthy relationships (Stanley, Amato, Johnson, & Markman, 2006). Efficacy studies on these early initiatives determined that RE resulted in improved problem-solving skills and relationship satisfaction, as well as reduced conflict (Carroll & Doherty, 2003). More recent meta-analyses have confirmed these early findings, both immediately after the intervention and in long-term follow-ups, indicating that RE programs produce moderate effect sizes on communication skills and relationship quality
There have been recent efforts to apply RE to individuals, often when psychosocial stressors preclude both partners from being present (e.g., due to childcare) or when safety issues exist within the relationship. Since 2011, there have been a number of RE initiatives for
individuals that have been subject to empirical study, and their findings suggest that they similarly produce reductions in psychological distress and relationship violence, as well as improved communication and relationship quality (Antle et al., 2013; Carlson, Rappleyea, Daire, Harris, & Liu, 2015; Visvanathan, Richmond, Winder, & Koenck, 2015).
Not only has RE been shown to help individuals and couples cultivate greater stability and support in relationships (Markman & Rhoades, 2012), RE has also been shown to serve as a gateway to subsequent couple therapy (Williamson, Trail, Bradbury, & Karney, 2014). Indeed, RE initiatives are designed to help individuals understand the benefits of cultivating skills and know-how within their relationship and ultimately feel more motivated and comfortable while engaging in future help-seeking behaviors. In fact, individuals who engage in premarital RE are three times more likely to pursue couple therapy than those who do not receive RE (Williamson et al., 2014). This is an especially notable statistic given the challenges that exist in engaging men in therapy. Men are less likely than women to seek psychological treatment, often due to concerns about sharing their feelings (Markman & Rhoades, 2012). Of the men who do seek therapy, attrition rates are as elevated as 50% (DeMaris, 1989). Yet, research on the efficacy of RE programs, which are predominantly psychoeducational and skills-based, suggests that no significant gender differences exist in the benefits that men and women incur from RE (Hawkins et al., 2008). This finding is consistent with research demonstrating that men have better
(Levant, 1990). Therefore, RE-based interventions create spaces that may be better suited for men who seek information and skills to improve the dynamics within their relationship.
Core elements of RE interventions. Since early RE interventions, the use of RE has since been expanded to diverse populations, including those with and without relationship distress. Among these programs, there are a number of core, shared elements. Specifically, most interventions contain a psychoeducational component. Psychoeducation, primarily intended to help patients and family members learn more about their illness, is an evidence-based
intervention for improving mental health outcomes that serves as a foundational component of most psychological treatments (Dixon et al., 2001). In addition to information about the disorder, psychoeducation modules may include skills development, as well as an introduction to
treatment. Psychoeducation has been used as a supplement to psychopharmacological
interventions across different forms of psychopathology including major depressive disorder, OCD, and bipolar disorder (Miklowitz & Johnson, 2009; Van Noppen, 2002). Across these disorders, the goals of psychoeducation are to maximize outcomes through engaging family members in treatment and improving recognition of symptoms (Lopez, et al., 2005).
In RE interventions, psychoeducation is designed to offer participants foundational relationship information (e.g., introduce the ways in which people may hold different standards within relationships) and couple-focused skills (e.g., communication skills; Halford &
In addition to psychoeducation, RE interventions are designed to improve communication and problem solving skills (Gottman & Silver, 1999). In order to facilitate this process, most RE initiatives seek to cultivate active listening skills (Gottman & Silver, 1999). Although
communication skills may be effectively implemented when both partners are present, there is demonstrated value in conducting RE for individuals (Bennun, 1997; Rhoades & Stanley, 2009). For example, relational conflict is associated with patterns of negative reciprocity (i.e., a
dynamic where one partner’s negative behavior increases the likelihood that the other partner will respond in a negative way; Salazar, 2015); however, by using active listening skills and speaking about one’s own opinions subjectively, that conflict is likely to take a different, less corrosive course as one partner learns new, more constructive communication patterns.
In the perinatal period, the presence of positive communication has been shown to protect against the development of mood disorders, whereas negative communication patterns serve as significant risk factors (Pilkington et al., 2015). Research shows that men have difficulty discussing emotions while also showing a willingness to discuss issues involving problem-solving or task-completion (Robertson, Grace, Wallington, & Stewart, 2004), highlighting the importance of active listening skills in this context. During the perinatal period, a time
characterized by fatigue and uncertainty, a partner who responds to distress with
problem-solving may inadvertently invalidate the patient’s experience and leave her feeling disconnected. In this way, teaching men how to cultivate and implement active listening skills can serve to foster support and understanding.
Relationship Education for Individual Partners of Women Hospitalized for PD
experiencing. For that reason, a partner-focused, individual RE intervention appears to represent a viable alternative. Although individual RE has not been applied to perinatal populations, there have been a host of RE group interventions for couples in the antenatal period designed to prepare couples for parenthood by promoting supportive behaviors, relationship skills (e.g., communication, conflict resolution) as well as providing psychoeducation about parenthood (Pinquart & Teubert, 2010). Results from a review of 21 couple-based RE programs for first-time parents suggested that the interventions generally did not produce improvements in communication nor in relationship satisfaction above and beyond minimal intervention groups (Pinquart & Teubert, 2010). More recent research, however, suggests that these types of
interventions may demonstrate greater efficacy for higher risk couples (Doss & Rhoades, 2017). Specifically, a number of recent RE studies revealed significant relational improvements among couples who possess factors that put them at risk for relationship distress during this time, such as financial stress or an unplanned pregnancy (Doss, Cicila, Hsueh, Morrison, & Carhart, 2014). Of course, psychiatric emergencies in the perinatal period also represent a risk factor. These findings suggest that while RE initiatives may not necessarily hold great utility across all new parents, they appear to be helpful for at-risk couples, including those enduring a maternal psychiatric emergency. Specifically, based on these RE results, as well as the literature on couples in the perinatal period, equipping partners with greater understanding and skills for offering support is likely to yield benefits for individuals and their relationships, both during and after hospitalization.
An Intervention for Partners during the Perinatal Period
many important benefits. Specifically, it could simultaneously reduce partner confusion and isolation during the patient’s hospitalization while also integrating men more into treatment. Involving men more in treatment would likely also result in improved support-giving behaviors. From a treatment continuity perspective, involved partners could serve as surrogate case
managers, reminding patients about medication compliance, future treatment appointments, and all of the other important components within the patient’s discharge plan. As noted earlier, a significant number of psychiatric patients do not follow-up with outpatient providers after discharge from inpatient units, and many others are not compliant with their medication (Olfson et al., 2000). Integrating partners would help to empower them to offer support, encouragement, and logistical help during a time when patients are often feeling challenged to fully re-engage with their lives. Indeed, the research on partner-assisted interventions for psychopathology, which are designed to improve the patient experience through facilitating partner support
behaviors, suggests that this is an effective avenue for improving patient psychological outcomes (Fischer et al., 2016). In that way, for reasons explored earlier, although improving partners’ ability to help patients make behavioral changes will likely aid in their immediate and long-term recovery, a couple-based intervention may not be well-suited for this specific population.
As such, the proposed partner-focused intervention was designed to be delivered during a single 90-minute session and include the following: psychoeducation about the patient’s
disorder, an exploration of the dyad’s current approach to living with the disorder, an
CHAPTER 2: CURRENT INVESTIGATION:
The current investigation had two primary aims. The first aim was to conduct a feasibility study of a unique intervention for male partners of women who are hospitalized for PD.
Feasibility studies are the recommended first step at the outset of intervention implementation and as such, this study did not utilize a comparison control group (Arain, Campbell, Cooper, & Lancaster, 2010). Recent guidelines put forth by Orsmond and Cohn (2015) suggest that
feasibility studies should address several key objectives. First, feasibility studies should examine recruitment capability, with a focus on the demographics of not only the participating sample, but also those who are ineligible or choose not to participate. To do so, the following feasibility themes were examined: ease of identifying patients and partners, diversity of the subjects, clarity of the eligibility criteria (e.g., is it clear who meets requirements for eligibility?), and refusal rate. Second, feasibility studies should be a means for clarifying data collection procedures, as well as the appropriateness of measures themselves, in order to refine these processes. Third, it is
important for feasibility studies to include measures of acceptability. Treatment acceptability is a term used to define the judgments about a treatment by its consumers (Kazdin, 1980). The standard in the field for measuring acceptability is to assess self-report participant satisfaction with the treatment through quantitative and qualitative methods (Kazdin, 1980). Fourth,
feasibility studies are designed to evaluate the resources available in a given treatment setting for implementing the intervention. The final objective of feasibility studies, per Orsmond and
intervention. As such, the secondary aim of this study is to examine preliminary findings relating to the intervention’s effectiveness.
Broadly, the intervention is designed to provide male partners with information about the patient’s disorder, skills for offering support, as well as a space for them to cultivate their own coping skills. As discussed earlier, psychoeducational interventions that improve partner understanding have been shown to reduce EE and improve support giving behaviors (Van Gent & Zwart, 1991). Within this intervention, male partners are also taught active listening skills to improve positive communication within the couple. In that way, the first two modules of the intervention are designed to improve the transmission of emotional support from partner to patient, given that perceived support represents an important resilience factor for mothers as it pertains to the development and maintenance of PD (Pilkington et al., 2015). Although positive instrumental support also promotes relationship health (Epstein & Baucom, 2002), research shows that women’s desire for emotional support often goes unmet within their intimate relationships (Xu & Burleson, 2001), and as such, it represents the main target of this intervention. Finally, partners explore different self-coping strategies to identify methods of promoting their own psychological well-being.
This is an empirically informed initiative that draws from the foundations of individual RE to deliver a single-session, partner-focused intervention for men whose wives are
hours). The majority of RE programs are designed to be delivered in the moderate-dosage range (9-20 hours; Hawkins et al., 2008). While the moderate-dosage programs tend to yield the greatest improvements in relationship quality and communication skills (Hawkins et al., 2008), there is evidence to suggest that low-dosage programs can also offer positive effects. For example, Braithwaite and Fincham (2007) implemented a one-hour, online RE intervention for individuals who were in romantic relationships. Their intervention was designed to help
participants cultivate communication skills and problem-solving techniques while also teaching them how to develop greater awareness for relational risk factors that may lead to relationship distress. The intervention resulted in significantly reduced individual depression and anxiety symptoms, as well as improved relationship quality at an eight-week follow-up as compared to the control, non-intervention group.
The proposed initiative is designed to make several unique contributions to the literature on treating women hospitalized for PD. Primarily, in spite of the fact that most women in the perinatal period have partners (Halford, Petch, & Creedy, 2010), severe PD tends to be treated as an individual disorder (Pearlstein, 2008). Yet, a lack of social support is one of the strongest, most consistent non-biological predictors of PD (Robertson et al., 2004) and although individual RE programs exist for couples transitioning to parenthood, no such interventions are designed for partners of women experiencing heightened levels of PD. As such, the goal of the current study is to assess the feasibility, acceptability, and effectiveness of the current intervention. As more attention and resources are allocated to addressing maternal mental health in the perinatal period, the proposed intervention sheds light on the benefits of partner engagement, both during
Hypotheses
Aim 1: To develop a partner-focused intervention and assess its feasibility and acceptability.
This intervention is designed to recruit partners and engage them in an individual
psychoeducational session to capitalize on them as a resource to aid in the patient’s recovery.
Hypothesis 1.1. Recruiting partners to be involved in this intervention will be feasible.
Hypothesis 1.2. Partners will find each component of the intervention to be acceptable.
Aim 2: To conduct a preliminary assessment of the intervention’s efficacy.
This one-session partner-based intervention is designed to improve understanding, support
behaviors, and coping skills among men whose partners are hospitalized for PD. This initial
evaluation will occur via an open-trial whereeach partner will receive the treatment, an approach
that is consistent with past research on interpersonal interventions. Specifically, in guidelines put
forth by Christensen, Baucom, Vu, and Stanton (2005) in their article on delivering
methodologically-sound interpersonal research, the authors advocate for open clinical trials in
order both to establish initial effectiveness and allow for an opportunity to make adjustments to
the intervention before moving forward with a more expansive randomized control trial.
Hypothesis 2.1. Partners will report significantly improved support self-efficacy
immediately after the intervention and at the four-week follow-up.
Hypothesis 2.2. Partners will report significantly reduced levels of expressed emotion at the
four-week follow-up as well as significantly improved relationship satisfaction and
Hypothesis 2.3. Patients on the unit will report significantly improved relationship
satisfaction, greater perceived support, as well as improved psychological well-being at the
CHAPTER 3: METHOD
Cultivating and Integrating Stakeholders
In the early stages of the intervention, we collaborated with the medical leadership team at the UNC Center for Women’s Mood Disorders. After proposing the intervention to the director of the UNC Perinatal Psychiatry Program, we presented the project at the monthly perinatal psychiatry research meeting to discuss the intervention, gain feedback from members of the team, and clarify the role of the proposed program on the unit where it would take place. Several other meetings occurred on the unit itself where the rationale for the intervention was discussed with nurses, medical students, and other members of the team to learn about the project and offer suggestions to help shape the intervention.
Participants
Twenty male partners of women in the perinatal period who were hospitalized on the UNC Perinatal Psychiatry Inpatient Unit (PPIU) were recruited to participate. For partners to be eligible for the study, patients had to first give consent, which allowed the study clinician to contact partners and discuss patients’ experiences during the intervention sessions. Partners of patients who were unable to provide informed consent (e.g., due to acute psychosis) were not eligible to participate. To be eligible for the study, a partner had to meet the following criteria: (a) at least 18 years old, (b) English-speaking, (c) a member of their current romantic
Measures
The measures in the study, completed by both patients and partners, were used to assess functioning across three domains: (a) partner response to psychopathology, (b) relationship functioning, and (c) individual psychological well-being. Qualitative and qualitative data were also collected to assess for feasibility and acceptability of the intervention.
Demographics (patient and partner). Both partners completed a brief demographics questionnaire assessing their age, ethnic background, gender, level of education, occupation, yearly income, treatment status (i.e., whether they are seeing a mental health professional), and length of romantic involvement with their current partner.
Acceptability (partner). In order to assess for acceptability of the intervention, partners completed the Client Satisfaction Questionnaire-8 (CSQ-8; Larsen, Attkisson, Hargreaves, & Nguyen, 1979). The CSQ-8 is an 8-item measure of client satisfaction that assesses for both the quality and type of service offered. Each item is rated on a 4-point, Likert-scale, where higher scores indicate greater client satisfaction. The measure has good internal consistently (Larsen et al., 1979) as well as strong construct validity (Attkisson & Zwick, 1982). Although there is no cut-off score to indicate acceptability on the CSQ-8, research from diverse samples, both in terms of demographics and services offered, have established norms from the CSQ-8 (M = 27.09, SD = 4.01; Nguyen, Attkisson, & Stegner, 1983), offering a basis for comparison. Several
open-ended questions were also used to assess for acceptability within each component of the intervention.
also collected to understand obstacles to recruitment. There are not currently any norms in the literature for recruitment rates for feasibility studies. As such, we identified six recent studies that targeted perinatal distress (Table 14) and six studies that utilized single-session interventions (Table 15) and calculated their respective recruitment rates in order to have a basis for
comparison.
Domain 1: Partner Response to Psychopathology.
Expressed Emotion (partner). Expressed emotion was assessed using the Patient
Rejection Scale (PRS; Kreisman, Simmens, & Joy, 1979), an 11-item questionnaire designed to capture criticism that family members direct toward individuals with psychopathology. Family members rate each item from 1 (never) to 3 (often); scores range from 11-33. The PRS has demonstrated high internal consistency (r = 0.78) and test-retest reliability (r = 0.72; Kriesman et al., 1979).
Support-Self-Efficacy (partner). Partner support self-efficacy (PSSE; Lorig, Chastain,
2008). Within each modified version, specific items have been added and removed to
individualize the scale for the specific disorder. For this study, the items have been adjusted to reflect the partner’s perceived self-efficacy in offering support for psychological distress, rather than pain, which is how it was originally written (e.g., ‘‘How certain are you that you can help the patient decrease her pain quite a bit?’’ was changed to ‘‘How certain are you that you can help the patient decrease her psychological distress quite a bit?’’).
Domain 2: Relational Functioning.
Relationship Satisfaction (patient and partner). Relationship satisfaction was measured
using the 16-item Couples Satisfaction Index (CSI-16; Funk & Rogge, 2007). Each item on the CSI-16 is rated on a 0-5 scale, aside from one item rated 0-6. To calculate a satisfaction score, scores are summed; higher score suggest higher relationship satisfaction. Examples of questions on the measure include, “Our relationship is strong” and “How well does your partner meet your needs?” The CSI-16 has demonstrated high internal reliability (α =.98) and construct validity that strongly correlates with other empirically-validates measures of relationship satisfaction (e.g., Dyadic Adjustment Scale, r=.91; Funk & Rogge, 2007).
Perceived Social Support (patient). The Support in Intimate Relationships Ratings
Scale-Revised (SIRRS-R; Dehle, Larsen, & Landers, 2001) was used to assess for social support in the patient’s intimate relationship. Specifically, the SIRRS-R is a 20-item measure, adapted from the original 48-item item measure. Within the measure, individuals are asked to rate how frequently their partner has performed various social support behaviors in the past month and how satisfied they are with that support. Both frequency and satisfaction are measured on 5-point Likert scales. Psychometric research on the original measure, the SIRRS, demonstrates strong internal