Psychopharmacology Decision-Making Among Pregnant and Postpartum Women and Health Providers: Informing Compassionate and Collaborative Care Women's Health

Abstract

Psychopharmaceutical use by pregnant and postpartum women is complicated by the complexity of prescribing as well as the sociocultural context in which medication-related decisions are made. This study sought to advance understanding of decision-making processes and communication experiences regarding use of psychopharmaceuticals during pregnancy by considering both provider and consumer perspectives. An electronic survey was conducted with health care providers (N = 88) and women consumers (N = 83) from July 2010 through October 2011 regarding the perceived costs and benefits of taking mental health medication during and around the time of pregnancy. Descriptive analysis compared and contrasted experiences between the two groups regarding consumer-provider communication, critical incidents and triggers in decision-making, and response to case scenarios crafted around hypothetical client experiences. Both similarities and differences were evident among health care provider and women consumer responses regarding costs, benefits, communication experiences, and case scenario responses. Both quantitative and qualitative survey results indicated the need for more accurate, unbiased, and complete information exchange around mental health and medication. Study results suggested the centrality of the client-provider milieu to guide decision-making and emphasized the expressed need within both groups to create a shared decision-making practice environment characterized by authenticity, non-judgmental decision-making, compassion, humaneness, and reciprocity.

Keywords:

• depression
• mental health
• pregnancy
• psychosocial
• postpartum
• beliefs
• anxiety

INTRODUCTION

Depression and other mental health disorders that occur in women during and around the time of pregnancy present a unique challenge to both medical and community-based systems of care which support women's health and well-being. The use of psychopharmacologic medication with pregnant and postpartum women is more complicated than in the general population due to the simultaneous consideration of both maternal and fetal/infant health and well-being, identified teratogenic effects of some medications, as well as the controversial sociocultural context in which medication-related decisions are made.

The past decade of epidemiological research has detailed the incidence and prevalence of depression, anxiety, and posttraumatic stress disorder during and around the time of pregnancy (Le Strat, Dubertret, & Le Foll, 2011; Czarnocka & Slade, 2000; Swanson et al., 2011). However, an expanded understanding of decision-making influences and service utilization options that emerge within the primary patient-provider relationship are necessary for advancing collaborative and compassionate care options for women. The patient-provider relationship is influenced by the fact that for many women, specialized perinatal mental health/psychiatric services may be unavailable, inaccessible, or unaffordable (Miranda et al., 2003; Rosen, Warner, & Tolman, 2006; Song, Sands, & Wong, 2004). As a result, obstetrics/gynecology providers—not psychiatrists or other trained mental health professionals—often play a prominent role in formal mental health assessment and treatment, including prescription of psychotropic medications (Peindl, Wisner, & Hanusa, 2004; Goodman & Tyer-Viola, 2010; Leddy et al., 2011).

Prevailing medical advice has focused extensively on the potential impact of psychopharmacology on fetal development and transmission of medications from mother to infant during lactation (Gentile, 2005; Jain & Lacy, 2005; Eberhard-Gran, 2005). This clinical advice has an empirical basis, although expanded research is needed to address the gaps in knowledge in this important area central to women's health (Mattison & Zajicek, 2006). Research reviews and consensus guidelines summarize these empirical findings regarding safety and efficacy to assist health care practitioners meet identified mental health needs emerging in women (Jain & Lacy, 2005; Altshuler et al., 2001). A recent meta-analysis conducted by Grote et al. (2010) further reinforces the assertion that untreated depression during pregnancy elevates the risk of low birth weight and prematurity. Consequently, many physicians conclude from existing research that these medications can be safely prescribed if carefully monitored (Nonacs & Cohen, 2003; Miller et al., 2008).

Simultaneously, a myriad of public consumer-focused materials and websites raise concerns regarding the safety and efficacy of medication use during pregnancy or lactation. These include books focused on potential safety concerns (e.g., Hale's (2006) “Medications and Mother's Milk”), as well as critical “anti-psychiatry” perspectives which raise concerns regarding biomedical bias leading to over-diagnosis and over-prescribing of mental health medications to advance pharmaceutical company interests. For example, Healy's (2012) “Pharmageddon” raises specific concerns and presents case situations regarding anecdotally harmful medication use during pregnancy. Women consumers may voice distinct views and preferences regarding psychiatric medications (Boath, Bradley, & Henshaw, 2004), which may be drawn from rigorous research studies, popular press literature, as well as personal and family experiences.

While technological advances have increased the wide-spread availability of resources and evidence-based information, little is known about the real world decision-making processes navigated by women consumers and their health care providers around whether to initiate and/or continue psychopharmacology during pregnancy and the postpartum period. Several studies have provided detail regarding psychopharmacology choices and preferences in clinical samples (Boath et al., 2004; Pearlstein et al., 2006; Battle et al., 2007). However, at this time researchers simply do not fully understand the processes health care providers use to weigh existing scientific information themselves, or when and how this knowledge is shared with consumers. Concomitantly, researchers know little about how women consumers weigh and consider information, personal beliefs, life experiences, and perceptions of their mental health needs and treatment preferences to make decisions about their use of psychiatric medication.

Preliminary Studies Informing Survey

The authors' joint work in this area began in 2008 through a preliminary study of semi-structured key informant interviews with women's health care providers and low-income women who had engaged in recent decision-making regarding psychiatric medication while pregnant or breastfeeding. The key informant interview questions asked physicians to detail a “best case” and “worst case” prescribing scenario from their practice. Similarly, researchers asked women to discuss their individual experiences related to psychopharmacology decision-making and to describe the key factors, positive or negative, which guided their decisions. Themes derived from qualitative analysis of these interviews included the importance of perceived costs and benefits, communication patterns that facilitated or hindered consumer-provider dialogue, and the importance of critical incidents or triggers that seem to effect collaboration or decisions to use or prescribe medication. These themes (and specific wording content) were used to develop a preliminary version of the survey data collection instrument, which was then subject to formal field testing study with an expanded group of health care providers attending at a national conference, as well as with pregnant/postpartum women receiving services at local community-based agencies. Specific feedback from these field tests and early Institutional Review Board (IRB)-approved research helped clarify wording, scope, and focus in the final version of the national data collection survey, increasing its face and content validity within the target sample population.

Current Study Objective and Research Questions

The primary objective of the current study was to deepen the understanding of the decision-making processes and communication experiences among women consumers and ob-gyn health care providers. Specifically, researchers addressed three research questions: (1) Do health care providers (HCP) and women consumers (WC) rate the relative importance of specific costs and benefits of taking mental health medications similarly; (2) What are the most frequently reported positive and negative communication experiences reported in both groups concerning consumer/provider communication; and (3) What are the critical incidents and triggers from both HCPs and WCs that have influenced their medication recommendations?

METHODS

Study Participants

The study was designed to sample purposively HCPs in obstetrics/gynecology who were well-acquainted with current guidelines for practice, and WCs who had recently engaged in a discussion and/or decision-making process regarding psychiatric medication use with a primary HCP. Inclusion criteria for the provider sample were a prescribing provider of health care to pregnant and/or postpartum women and currently affiliated with an academic medical setting. Inclusion criteria for WCs were being currently pregnant or pregnant within the past two years; over age 18 years; and having had a discussion with an HCP regarding mental health medication during pregnancy or postpartum, regardless of whether she ultimately decided to use psychopharmaceuticals. Individuals selected for inclusion in the study were requested to affirm that they met inclusion criteria (so that eligibility was not independently confirmed by study staff) and had read the informed consent information before proceeding to the survey. The IRB at Virginia Commonwealth University reviewed and approved the study protocol.

HCPs were recruited from a sampling frame of clinical women's health providers affiliated with academic medical centers. This sampling frame was chosen to reflect practitioners most likely to be aware of and engage in current, evidence-based practices. Researchers initially generated a list of 131 academically-affiliated medical centers within the United States. A total of 50 institutions were selected to have one representing each state; in states with more than one center, researchers selected an institution at random from those listed. From within the 50 selected academic medical centers, researchers identified professionals with a prescribing degree (MD, DO, NP, PA, or CNM). Six contacts were selected from each institution using a randomization scheme based on institutional size. Names and e-mail addresses from the website were used for recruitment, but no connection was made between the name sent in the invitation and participant response, so that response rates could not be determined. A reminder e-mail was sent at 7, 14, and 30 days following initial contact, and participants could forward the survey link to national colleagues. This yielded a final sample of 88 HCPs (29% response rate).

WC recruitment was guided by respect for privacy and confidentiality. Researchers used a direct marketing approach as an effective and respectful way to recruit participants who met study criteria but who could elect participation and maintain their confidentiality. First, researchers targeted existing Internet sites where pregnant and postpartum women sought information. They included both paid and gratis web link insertion (with IRB approval number) on numerous websites including non-profit perinatal mental health organizations, non-profit parenting support organizations, as well the largest Internet provider of parenting information, Parenting.com. Second, researchers actively recruited via social network media. They established Facebook and YouTube sites which contained direct links to the survey recruitment website. To generate social media interest, they linked their site to other Facebook pages of interest to pregnant and postpartum women, as well as to mental health consumers. Third, they engaged in direct marketing using a brief recruitment message which was distributed widely to perinatal mental health clinicians via listserve announcements to announce/post openly within their practice settings. A total of 199 surveys were initiated. However, the survey could not be completed if participants did not affirm meeting inclusion criteria. After engaging in all three recruitment strategies over a 16-month period, a final sample of 83 women who met all inclusion criteria completed the study (42% eligibility/response rate).

Data Collection and Instruments

The mixed-methods data collection instruments used in this study are available for viewing at http://wp.vcu.edu/skprice/. This website contained an active link to initiate the survey during data collection. Survey contents reflected perception, personal experience, as well as participant guidance about “next steps” in two case studies which were presented similarly to both groups. One case study (“Carla”) described a first-time mother in the postpartum period who revealed a recent onset of depressive symptoms and who took an antidepressant medication in the past. The other case study (“Barbara”) described a woman who subsequently became pregnant following a stillbirth and who was experiencing anxiety symptoms in the current pregnancy. The survey data collection was active from July 2010 through October 2011.

Data Analysis

Researchers used a mixed-methods approach to data analysis, concurrently examining broad trends and specific nuances of the survey data. Quantitative analysis was conducted using IBM SPSS Statistics 19, descriptively examining patterns of response and observed trends in responses which were distinct to HCPs and consumers. Research questions were descriptive rather than analytically predictive; therefore, researchers only examined bivariate differences between the WC and HCP groups using independent sample t-tests and/or chi-square tests, appropriate to the type of measurement, continuous or categorical, respectively, but did not pursue additional multivariable, inferential analyses.

In the qualitative analysis, researchers employed manual coding of open-ended responses to questions which addressed the research questions: costs and benefits, decisional influences, and communication patterns. They employed a traditional thematic analysis within each of those areas, which is consistent with the positivist orientation and the overall mixed-methodology approach. The qualitative analysis involved an iterative process of data reduction, moving relevant text from the raw data into identified and organized themes which mirrored the research questions. The themes were examined first in relation to each participant group independently, and then the themes were interrogated to draw nuanced contrasts and comparisons. To promote quality control, the authors engaged in a collaborative data corroboration process whereby the first author, who had independently conducted the quantitative analysis, and the second author, who had independently conducted the qualitative analysis, audited each other's respective findings. When agreement was found, they incorporated that element into the study findings. When disagreement was noted, they reviewed the audit trail and raw data and reached consensus on whether or not the finding was conclusive. This approach to mixed-methods analysis has been described in detail by Bazeley (2009). At the conclusion of the study, researchers consulted with an outside expert in qualitative analysis who confirmed the rigor of the process.

RESULTS

Survey data were obtained from a total of 171 respondents, 88 HCPs and 83 pregnant and postpartum WC. The mean age for HCPs was 47 years (SD = 10.3), and 87% of these respondents self-identified as white/Caucasian, 9% as black/African-American, 3% as Native American, and 1% as Hispanic/Latina (Table 1). HCPs were predominantly women (65%), and 75% were practicing MDs in obstetrics/gynecology, while 9% identified as nurse practitioners, and 16% did not report. Practice locations were urban (62%), suburban (32%), and rural (7%). The average number of years reported in practice was 16 with the largest proportion (30%) reporting having been in practice over 20 years.

TABLE 1 Demographic Composition of WC and HCP Groups

	WCs (N = 83)	HCPs (N = 88)
Age, years (mean)	34 (SD 5.9)	47 (SD 10.3)
Race:
White/Caucasian	90%	87%
Black/African-American	2%	9%
Asian/Indian	3%	3%
Native American	3%	N/A
Hispanic/Latina	2%	1%
Education:
High school/GED	6%
Some college	10%
College degree	37%
Graduate/Professional degree	47%
Currently pregnant	18%
Currently postpartum	10%
Experienced pregnancy loss	39%
Experienced stillbirth/infant death	2%
Currently taking psychiatric meds	59%
Taken psychiatric meds in past	74%
Provider type:
Obstetrician/Gynecologist		75%
Nurse practitioner		9%
Other/Not reported		16%
Practice location
Urban		62%
Suburban		32%
Rural		7%
Years in practice (mean)		16 (SD 9.7)
Less than 5		12%
5–10		24%
11–20		34%
Over 20		30%

The mean age of the WC participants was 34 years (SD = 5.9). Race and ethnicity were also predominantly white/Caucasian (90%), with only 2% black/African-American, 3% Asian/Indian, 3% Native American, and 2% Hispanic/Latina. The education level of participants was skewed toward highly-educated women: Forty-seven percent reported a graduate/professional degree and 37% a college degree. A total of 18% of WCs responding were currently pregnant, 10% were currently postpartum, 39% had experienced a prior pregnancy loss, and 2% had experienced a stillbirth or infant death; 59% were currently taking psychiatric medication, and 74% had taken psychiatric medication at some point in the past.

Perceived Costs and Benefits

The first overarching question in the analysis examined the perceived importance and weighting of the range of specific costs and benefits of taking psychiatric medication reported by WC and HCP participants. The groups agreed regarding several important benefits of psychiatric medication (Table 2). Statistically-significant differences largely reflected nuances between items considered “very” versus “somewhat” important to decision-making between groups. Notably, no HCPs ranked concern about relapse or suicide in their patients as “not at all” important. The benefit of helping to deal with loss and bereavement reflected a greater degree of difference between the “very important” and “not at all important” categories for women, perhaps reflecting a degree of personal relevance of specific items on the part of women respondents. A relatively low degree of concern was observed related to stigma and negative self-identity by both groups, with surprisingly few women rating either of these perceived risks “very important.”

TABLE 2 Perceived Benefits and Costs of Psychopharmacological Medication Use Among WCs and HCPs

	WCs (N = 83)			HCPs (N = 88)
	Very important	Somewhat important	Not at all	Very important	Somewhat important	Not at all
BENEFITS
Helping feel more normal	43%	48%	9%	40%	44%	16%
Helping get sleep or needed rest	45%	40%	15%	52%	40%	8%
*Improving daily functioning	65%	28%	7%	86%	13%	1%
*Helping deal with loss and bereavement	50%	34%	16%	63%	35%	2%
Helping reduce stress around parenting	47%	37%	16%	35%	55%	10%
Supporting better bonding with baby	63%	30%	7%	63%	32%	3%
Supporting better family relationships	49%	37%	13%	53%	53%	4%
*Reducing risk of relapse	55%	37%	7%	87%	13%	0
*Reducing risk of suicide	75%	16%	10%	94%	6%	0
*Reducing mental health symptoms	64%	34%	2%	81%	18%	1%
COSTS
**Concerns about harm to fetus	89%	9%	2%	66%	30%	4%
**Concerns about infant development	88%	10%	2%	56%	36%	8%
**Concerns about meds in breast milk	68%	28%	4%	25%	58%	17%
*Concerns about medication dependence	28%	42%	30%	13%	43%	44%
**Concerns about long-term child effect	82%	16%	3%	48%	35%	16%
**Perception of stigma and judgment by others	5%	33%	62%	13%	47%	40%
**Negative impact on self and identity	5%	51%	44%	25%	52%	23%
Experiencing unwanted side effects	22%	49%	29%	26%	58%	16%
*χ^2 significant at <.05; **χ^2 significant at <.005.

Both WCs and HCPs were asked to note which benefits seemed to have shifted in importance for them over time. Among the women, two predominant themes emerged from the analysis of narrative responses to this question. Achieving normalcy in daily life: Women discussed the growing importance of being and feeling like “oneself,” being able to do everyday tasks more easily/willingly (dressing and self-care, preparing meals, getting to work, and being productive), as well as medication helping them to get sufficient rest and adequate sleep. Sustaining the family: Women discussed potential benefits of medication use, including the growing importance of bonding with their baby as well as being able to fulfill their roles as both caregivers and companions. A reduction of stress was described by many as foundational for achieving a satisfactory quality of life for themselves, their baby, and their entire family. Importantly, HCPs reported very similar shifts in thinking about benefits, namely the growing importance of psychosocial functioning and quality of life as a consideration in decision-making about psychiatric medication.

The data also revealed differences between groups related to costs and benefits. First, while both WCs and HCPs continued to be concerned about the impact of psychiatric medication on fetal, infant, and child development, these perceived risks rose to the top for WC participants with over 80% endorsing these items as “very important” risks. For HCPs, the risk was split between “very” and “somewhat” important for each of these three items. Concerns about medication in breast milk were more often reported as “somewhat” important for HCPs while a majority of women rated this concern as “very important” as a cost in weighing their medication decisions. While 62% of women responded that stigma and judgment by others was “not at all” important as a cost in deciding to take or not take psychiatric medication, HCPs appeared more concerned than WCs; a statistically significant difference was observed between WCs and HCPs overall regarding both the importance of social stigma and negative impact on self/identity as potential costs of psychiatric medication use.

Consumer-Provider Communication

A descriptive rank-order comparison was conducted regarding which declarative exemplar statements most closely matched their recent experiences with consumer-provider communication concerning psychopharmacology (Table 3). The survey items quoted actual statements made during key informant interviews in the earlier phase of this research. Interestingly, 13% of HCPs and 27% of WCs did not endorse any positive statements. A sizable 42% of both HCPs and WCs did not endorse any negative statements. Overall, fewer positive items were endorsed by WCs than HCPs (t = –2.23, p = .027; mean difference between groups –.41, 95% CI –.772, –.047), but the total number of negative items endorsed between groups was not significant.

TABLE 3 Rank Order Analysis of Statements Reflecting Consumer-Provider Communication

	Positive statements	Negative statements
WC rank:
#1	Paid attention to what I was saying/doing (31%)	I wasn't quite sure what to say (28%)
#2	Provider seemed to “get it” (25%)	I thought I had better keep some things to myself (27%)
#3	I had the opportunity to ask all the questions that I wanted (23%)	I felt I was not given a choice (21%)
#4	I felt understood (23%)	I felt rushed (13%)
#5	I felt respected (23%)	I felt ashamed (13%)
HCP rank:
#1	We talked openly about risks and benefits (71%)	Felt a bit beyond my expertise (31%)
#2	We both actively participated (52%)	Expected to prescribe without delving too deeply (25%)
#3	My professional experience was valued (47%)	Patient seemed to hold things back (22%)
#4	I felt it was a balanced discussion (38%)	Frustrated by the discussion (9%)
#5	Information was freely shared (24%)	Had doubts about the decision (8%)

Respondents were asked to choose one of the positive or negative statements from the provided lists and “tell us more” in narrative form about how it might relate to their own experiences with decision-making around psychiatric medication and pregnancy. HCPs took the opportunity to offer considerable reflections on the limitations and contextual challenges to the medical encounter with pregnant and postpartum women in need of significant emotional support, mental health care, and/or psychiatric medication. While the quantitative data already hinted at the frustrations of these mostly ob-gyn HCPs, the narrative data detailed more specific challenges, including: frustration with clients who do not follow up on their referrals for either psychosocial or psychopharmacological care; lack of depth in their psychopharmalogical training, both when trained originally and their current continuing medical education offerings; challenges in practicing in the information age where popular sources of information may be more influential with clients; lack of accessible local mental health and psychiatric resources leading to higher expectations for ob-gyn HCPs; and seeming to bear the brunt of responsibility for drawing out clients and engaging them in decision-making.

The quantitative data from WCs suggested that provider attentiveness was a most crucial and desirable experience for women. However, one of the most compelling findings in the narrative data about positive communication was the degree to which women endorsed the notion that provider openness—and specifically a lack of being judgmental—was essential for productive and effective communication. WCs said they were seeking respect for their own intelligence and experience and wanted providers to share their knowledge and scientific information with them. They also desired reassurance in a caring atmosphere, deeply valued when providers took action on their concerns, and importantly, followed up on those actions. Taking action emerged as important in the responses to the case scenarios as well.

The qualitative elaborations by women about negative communication and interactions with providers also echoed the quantitative data. The most prominent theme was the notion that women were keeping some things to themselves, and indeed sometimes even lying about their situations out of fear of being judged, being seen as not normal, or even out of fear of having their children taken away if they disclosed too much about their symptoms. Many women discussed being ashamed of their own depression and anxiety, having significant discomfort with talking about it, and moreover that they did not know how to talk about it. With respect to HCPs, women readily expressed frustration with feeling disrespected, dismissed, and belittled, as well as “being given a pill without much discussion” such as education or follow-up. Some women reported feeling that choices were never really offered nor discussed.

Response to Case Scenarios

Case #1: Carla

Carla was a 29-year-old woman who recently gave birth to her first child. It was a good birth experience overall. Indeed, she liked being pregnant and was able to breastfeed, although it was challenging at first. She took care of herself and mostly managed her occasional bouts with depression through exercise and relaxation. She took an antidepressant for a year some time ago when she was going through a particularly difficult time related to work and a personal relationship. She thought the medication was quite helpful. Since her baby was born, however, she has been feeling like she “is going through the motions” and has been less and less excited about motherhood. In fact, she was tired and lonely all the time and sometimes even wondered why she had the baby. She hasn't been caring much about her appearance, doesn't fit in her clothes, and feels unattractive. On many days, it was a chore for her to get out of bed. She was scheduled for her postpartum check-up with her ob-gyn provider today.

What did women and HCPs think was the next step in working with Carla? Would they as providers, or if they were Carla herself, broach the topic of prescribing medication? For women, five different themes emerged from a thematic analysis of these open-ended responses. Most predominant in terms of “next steps” was that women said providers should be drawing out Carla's feelings with direct questions about her emotions and state of mind. They offered very specific active listening phrases and specific questions that might be a way to begin to do that. In addition, they believed that providers should do what they could to normalize Carla's experiences and explicitly offer her hope and reassurance. They also suggested that Carla be offered advice and be connected to resources and information for support. One of the most interesting findings was that the women not only suggested what providers should be asking and talking about (motherhood, life), but often empathically noted how they should be doing it, that is, under what conditions. For example, not just ask but to ask “in a kind voice,” or “look at her right in the eyes,” or “have her sit down,” and make sure to talk “while she is dressed.”

Researchers then asked WC participants to say whether or not they would broach the topic of psychiatric medication if they were Carla. In their responses, 47% of women leaned toward encouraging Carla to discuss medication, while 30% discouraged discussion of medication. The remaining 23% of women remained uncertain or listed equally both pros and cons of initiating the conversation. The most compelling influence for medication use for women was the fact that Carla had successfully used medication in the past. Others reported such influences as her symptoms being “beyond the baby blues,” and “you can't take care of your baby if you don't take care of yourself,” or “she needs to enjoy motherhood.” They noted Carla needed medication so she could be “emotionally present to her child” and “not risk losing important bonding time.” One woman noted if she were Carla, she might benefit from antidepressant medication but would not actually seek medications and “admit defeat.” Several others noted that breastfeeding in this case was the deal breaker, and medication should be seen in those cases “as a last resort.” Those ambivalence about medication seemed to defer to the provider, letting her or him suggest medications or present other options like exercise and therapy.

Providers overwhelmingly noted in Carla's case that they would screen her for depression, suicide, and self-harm. HCPs leaned heavily toward recommending formal screening (47%), and for approximately 26% of providers, the discussion moved beyond screening to include a referral for mental health evaluation and/or a prescription, usually an antidepressant. Interestingly, in Carla's situation, only 4% of HCPs specifically mentioned referral and/or intervention based in psychosocial or behavioral support. When asked what would “trigger” a referral for medication, clear themes emerged from the responses. Predictably, they would refer if she had a high score on a depressive symptom screening inventory or if she had a marked impairment in ability to carry out the tasks of daily living. In spite of their subtlety, broad consensus was apparent among providers around three other very important triggers: a mother's lack of interest in her baby, a mother's isolation from the world, and an uncharacteristically deep sadness. Health providers frequently named two or more things which together would have to be present for them to broach the topic of medication. For example, it might be a high score on a screening test and expressed thoughts of harming oneself, or a past history of success with an selective serotonin reuptake inhibitor (SSRI) and a lack of bonding with the baby.

Case #2: Barbara

Barbara was a 33-year-old woman who has been going to the same ob-gyn provider for over 4 years. She had one child, age 4. Two years ago, she had a stillbirth that was devastating to her. She and her spouse waited a year to try to get pregnant again. After trying for approximately 9 months, they were successful and she was currently pregnant. Now, at 21 weeks into her pregnancy, she was becoming increasingly worried that “something is going to happen to this baby.” She frequently called her Ob-Gyn office to ask questions about cramping, spotting, and contractions. She wished she could have more ultrasounds to make sure everything is ok; she wondered when she should start “kick counts” to make sure the baby is moving. She was avoiding questions from family and friends, believing that she might “jinx” her current pregnancy. Barbara remembered a few times in the past when she felt very worried and took an anti-anxiety medication, and she wondered if she needed it again. Her regular prenatal visit with her provider was this afternoon.

Barbara's case scenario was based largely in situational depression and anxiety occurring during a subsequent pregnancy following stillbirth. While her symptoms were evident, the case provided details of life events that were potentially contributing to her challenges in well-being. What did women and HCPs respectively think was the most important thing to do next with Barbara? The short answer for the women respondents was several things. That is, most women offered not one, but a number of things providers should be doing with Barbara, the most prominent of which was to acknowledge directly Barbara's past perinatal loss and offer her reassurances and information that might allay her worries in that regard. Their suggestions were clear that while providers should discuss and explore her worries and validate her experiences, they should also act: educate, screen, offer support, direct to counseling, provide more frequent visits, and suggest exercise and stress reduction techniques. In their responses, 45% of women indicated they would likely bring up the subject of medication, 29% stated they would not, and 27% gave no clear indication of their response. It is noteworthy that the women who indicated they might address the topic did so in a “yes, but” or “it depends” fashion. For example, “Yes, but I would have mixed feelings;” “Yes, depending on how anxious she was,” “Perhaps, but I want to hear the pros and cons first;” or “Maybe, but only as a last resort.” An explicit discussion of the risk-benefit equation was prominent in WC responses to Barbara's case, no matter on which side of the medication decision they landed.

HCPs' responses echo similar assertions as WCs, noting that they would “ask,” “inquire,” “acknowledge,” or “explore” Barbara's experiences with anxiety and fear related to her past loss and offer reassurances and connections to support resources. Secondly, they would use facts and statistics as a basis for reassurance about the current pregnancy's progress. In Barbara's case, it was good news that what the women said they would want providers to do, the providers said they would do in this regard. HCPs and women participants were more reluctant to prescribe and instead preferred referral for counseling first. Thirty-eight percent of HCPs stated they would engage her in a general, open conversation (vs. 21% formally screening her for mental health symptoms), and 23% intended to refer or provide psychosocial and behavioral support. Only 9% would immediately consider prescribing a medication for her mental health symptoms. Critical incidents and triggers listed for prescribing included suicidal ideation, risk to self or others, and extreme anxiety which impaired daily functioning.

DISCUSSION

The study's cross-sectional design, purposive sampling methods, and modest sample size require caution when interpreting the results. Furthermore, both samples were composed of highly-educated and relatively homogeneous samples in terms of race and ethnicity and high educational attainment, and thus cannot be considered broadly representative of either population. HCPs represented providers linked with academic medical centers and thus did not represent the range of HCPs encountered in community practice settings. Similarly, the inclusion criteria for WCs with specific experience discussing mental health and medications with an HCP resulted in a sample of women that was not representative of the “average” pregnant woman. It is possible that both HCPs and WCs who responded to the study self-selected to participate based on interest in the topic, and/or that social acceptability guided their response to individual items on the survey. Furthermore, researchers did not confirm eligibility in either group. Also, participation bias was a possibility in this study, given the low participation rate from the HCP sampling frame. These sampling and design limitations could affect the accuracy and generalizability of results. Finally, although researchers used peer auditing and iterative processes of data interrogation to enhance rigor, they did not engage prospectively in independent coding of transcripts by multiple coders as a quality control measure in the qualitative analytic approach. These important limitations to generalizability and accuracy should be kept in mind when interpreting the findings. Therefore, the researchers present descriptive findings from their study which offer points of clarity and conundrum for the reader, relevant to designing future research and considering practice innovations.

Clarity

Both the risks of taking psychiatric medication, especially the known and unknown physical effects of medications on the fetus and infant, and the benefits, especially the significant positive impact on the well-being of the mother as well as the quality of family life, were similarly recognized by both women and providers. The relevance of these costs and benefits as factors influencing decision-making was consistently affirmed by both groups throughout the study.

With respect to decision-related communication between women and providers, a vast range of negative and positive experiences were presented. The creation of a caring and action-centered communication milieu was central to a foundation of shared decision-making and true dialogue, which both parties deeply desire. For providers especially, suicide, self-harm, and significant difficulties in daily living were triggers to broach the topic of psychiatric medication with pregnant and postpartum women. However, more subtle cues of lack of interest in the baby and isolation from the family also drew consensus as critical factors, especially when seen in combination with other concerns.

Conundrums

The nuanced differences between providers and women with respect to perceived safety and efficacy of psychiatric medication presented a conundrum in this study, with consumers expressing greater concern than health providers regarding the effects of medications on fetal, infant, and child development. This conundrum may relate to the sources of information used by the different groups, with providers presumably resting their professional perceptions based on clinical studies reported in medical journals, and WCs resting their decisions on anecdotal reports, personal experience, and publically available media reports of research. Additionally, HCPs brought a wider perspective to the conversation, viewing the differential experiences of many women within their professional care, which may have moderated the overall degree of concern over “what might happen” in any one specific situation. Women, understandably, were focused on the central importance of their own situation.

A conundrum with respect to communication was that women expressed frustration that sometimes medication is offered without much discussion, while at the same time providers expressed frustration that they are sometimes expected to prescribe without delving into detail about the patient experience. These seemingly incongruent findings may speak to the broader context of psychopharmacology practice and the hurried nature of clinical encounters in contemporary medical settings. It seemed to affirm the renewed interest in shared decision-making in both medical and psychiatric practice (Loh et al., 2007; Adams & Drake, 2006) and offered the opportunity to consider approaches that free up the conversation between women and health providers around this important topic.

The range of responses to the case scenarios suggested a larger conundrum, that opinions and responses and decisions as to “what should be done” vary greatly in real world practice. While this was consistent with emerging focus on patient-oriented outcomes and shared decision-making processes, it also created a conundrum with the real-world demand for “best practice” guidelines that are simultaneously responsive to a wide range of contextual factors and personal preferences. As researchers build scientific knowledge regarding the interrelationships between maternal depression and fetal outcomes, as well as the long-term effects of specific medications during pregnancy and lactation, the balance of risks and benefits may continue to fluctuate.

Finally, study findings indicated the relative unimportance of stigma surrounding mental health issues, specifically related to psychopharmacology decision-making. Nevertheless, the current data still showed that many women were ashamed of their depression, anxiety, and/or other mental health symptoms, had considerable discomfort talking about it, and indeed many said they did not know how to talk about it. Resolution to this particular conundrum may rest in opening up dialogue regarding depression and other mental health concerns that arise during and around the time of pregnancy in the broader, social arena so that the discussion of “depression” is not synonymous with “antidepressant medication.”

Both providers and consumers had important dialogical concerns related to what the best and most appropriate response should be to emergent or returning mental health challenges in pregnant and postpartum women. But the discussion cannot solely rest in the brief exchanges around mental health and medication which occur amid health care visits that are meant to address a wide range of human health concerns. The role of social work, nursing, and public health providers in opening up communication and linking both women and health providers with key resources for shared decision-making is an important next step to consider. Indeed, the study provides encouragement to formally investigate roles for multi-disciplinary health professionals including expanding dialogue, and testing research and practice innovation that begins to alleviate the felt pressures expressed by many HCPs of trying to be the “default mental health provider” for women encountered in routine practice.

Final summative thoughts are offered here which echo a recent web-based survey of 100 pregnant and postpartum women conducted by Patel and Wisner (2011), who also found, not surprisingly, most preferring “active” or “collaborative” or “shared” roles. The current study suggests that women were craving three specific things from providers: First, the participants wanted accurate, unbiased, and complete information on what is known and not known, normal and not-so-normal about mental health, medication, pregnancy, and post-pregnancy. Second, they wanted the milieu of the client-provider encounter to be a non-rushed interaction and dialogue characterized by caring, humaneness, and reciprocity. This interaction reinforces that they are not alone in their struggles, and that providers can tolerate the ambiguity inherent in decision-making and the weighing of risks and benefits. Third and finally, they wanted providers to act and take action: to educate, refer, screen, assess, listen, problem-solve, and address the many splendored dilemmas that they present with respect to their mental health. What is most encouraging and important about the study was the recognition that HCPs also wanted to practice in a contemporary context that allows all of that to occur. The challenge is to collaborate effectively around meaningful models of care and augmentations to existing services, which facilitate this shared aspiration.

Notes

The project described was supported by the VCU School of Social Work small grants program and the first author's KL2 award No. KL2TR000057 from the National Center for Advancing Translational Sciences. Neither author has any conflict of interest associated with this study or publication of its findings. Both authors contributed equally to the conceptualization, conduct, and analysis of the study as well as to writing and editing of the manuscript.