Enhancing support for postpartum psychosis through a socio-ecological lens: A qualitative analysis

Design

The Massachusetts General Hospital Postpartum Psychosis Project (MGHP3), launched in 2018, enrolls an international cohort of individuals who experienced PP within 6 months of childbirth and during the 10 years prior to study participation, contributing to a greater understanding of the etiology, presentation, and treatment of postpartum psychosis (Cohen et al., 2023). Participants are recruited primarily through internet presence, PP advocacy networks, and physician referral. A one-time structured clinical interview is conducted with participants by phone to confirm PP diagnosis and capture lifetime psychiatric and medical history, including postpartum treatment information. A detailed description of the MGHP3 methodology has been previously reported (Cohen et al., 2023). This manuscript reports on a cross-sectional self-administered follow-up qualitative survey that was administered to all enrolled MGHP3 participants. All study procedures were conducted following the Declaration of Helsinki and with approval from the Mass General Brigham Institutional Review Board (protocol #2018P000886, approved August 2018). Privacy rights of human subjects were observed, and informed consent was obtained from all participants.

Measures

The self-administered follow-up survey contained six short written questions (see Figure 1) about influences, factors, and changes that would have been beneficial prior to, during, and after their PP episode.

OPEN IN VIEWER

Figure 1.

Survey questions. The six original questions posed in the self-administered follow-up qualitative survey, with question 5 being the focus of this analysis.

This analysis is of open-text responses to one question, “How do you think those who experience postpartum psychosis can be best supported?” The questions in this survey were developed based on clinical interest of the research team, input from MGHP3 advocacy partners, and to build on a qualitative analysis of PP narrative experiences from this study sample (Vanderkruik et al., 2024). The survey questions did not change over the course of the data collection period.

Participants

Participants were eligible if they completed the MGHP3 interview before October 1, 2023. The first six-question follow-up survey invitation was sent out on December 1, 2023, with a reminder sent in January 2024. Data were collected up until February 10, 2024. For more information on the eligibility criteria for MGHP3, see Cohen et al. (2023). Participant demographics were collected in the initial MGHP3 interview before the self-administered survey. A $25 Amazon gift card was provided to individuals who completed the self-administered qualitative survey.

Analysis

A content analysis approach was applied to identify major and minor themes (Elo and Kyngäs, 2008). We employed a combination of inductive and deductive coding to analyze survey responses. Initially, researchers familiarized themselves with the data by thoroughly reading the de-identified responses. An inductive approach was used to identify emerging codes. As codes developed and were refined over the analysis process, the coders recognized that emerging codes aligned with the SEM framework (Fereday and Muir-Cochrane, 2006). In the analysis process, initial codes were reorganized to be child code of each section of the framework. We applied a version of the SEM that was adjusted to specifically address mental health and well-being to guide deductively generated codes; the model shifted the narrative from mental illness as an individual’s responsibility to the impacts of social and environmental influences on the well-being of those diagnosed with a mental illness (Golden and Wendel, 2020; Michaels et al., 2022).

Within this SEM theoretical framework, participant responses regarding types of support from those with lived experience of PP were mapped at each of the levels of influence. As outlined in section 1 (introduction), there are six levels of the SEM: the intrapersonal level, the interpersonal level, the organizational level, the community level, the policy level, and the societal level. Each level is measured with a frame that focuses on improving mental health (Erickson et al., 2024). The intrapersonal level is measured based on the relationship one has with themselves. The interpersonal level measures the relationship between people. The organizational level is measured by participant suggestions for clinical care changes that range in complexity. The community level is measured in the relationships which occur in diverse settings (e.g. support groups or support for those within the participant’s social network; Bronfenbrenner and Morris, 1998). The policy level is measured by looking at how the individual is impacted by employment and/or governmental policies. Finally, the societal level is measured by the impact and influence society has on the individual.

Data were deidentified, cleaned, and manually coded using Microsoft Excel. Coders first familiarized themselves with the data by reading through each of the participants’ responses. The first round of coding was then conducted by reading through each of the participants’ responses to develop initial codes, and the last two rounds of coding were conducted with responses grouped by question rather than by participant to ensure consistency in coding across content areas. This choice may have affected the contextual or biographical coherence of the data, however, after the first two read throughs, coders felt comfortable with the flow of the narratives. Initial code generation was conducted collaboratively by four researchers with varying levels of expertise, two holding Bachelor of Arts degrees (HEY, MG), one holding a Bachelor of Science degree (PB), and one holding a Master of Public Health degree (AS). This team met regularly to discuss the coding process, compare interpretations, and resolve discrepancies. Reflexivity is an important cornerstone of qualitative research (Berger, 2015); therefore, researchers regularly questioned the impact of their backgrounds and beliefs on the interpretation of data and generation of themes. Disagreements were resolved through discussion to better refine code definitions. These results were collated and interpreted qualitatively. A fifth researcher contributed to decisions in cases of disagreement (RV).

A codebook was developed and iteratively refined throughout the analysis to ensure clarity and consistency of codes. After initial coding, each researcher independently analyzed specific survey responses. To ensure reliability, two researchers reviewed the others’ coding work, verifying alignment with the codebook. Once coding was finalized, codes were grouped into preliminary themes, focusing on question five of the survey to begin organizing the data. The themes were refined based on their prevalence across survey responses, ensuring they represented cohesive and meaningful patterns within the data set. The data extracted supporting each theme were reviewed to confirm their relevance and consistency. Researchers collaboratively reviewed and refined the themes, defining and naming the final themes to accurately capture the core ideas within the data. The standards of reporting qualitative research (SRQR) were used to guide the analysis and reporting of findings (O’Brien et al., 2014).

Individual level

At the intrapersonal level, a key theme that emerged was the importance of self-acceptance and trust for recovery. A participant emphasized, “She needs to know that it’s not her fault” (P1328). Participants discussed how repairing one’s self-concept and the ability to forgive oneself and have improved self-awareness influenced their experience of and recovery from PP. They emphasized that knowing that recovery is possible can empower individuals facing PP. One participant shared, “Being honest with yourself about how you’re feeling and not doubt[ing] yourself. Don’t be ashamed to take care of yourself . . . ” (P1004). Participants also underscored the importance of accepting support from others, emphasizing that seeking help should not be viewed as shameful or as a sign of failure, but instead as an important part of self-care and recovery.

Interpersonal level

Four themes emerged at the interpersonal level:

Community level

At the community level, three themes arose:

Organizational level

At the organizational level, participants proposed three key areas for improvement:

OBGYNs need to be trained in warning signs and able to prescribe the medications needed to help with psychosis. Every doctor who wasn’t a [psychiatric health care provider] was terrified to prescribe me the anti-psychotics needed, and had I not gone to the ED and immediately been put in a program with access to a psych NP immediately, I would’ve had to wait 6 weeks for an appointment with a [psychiatrist] for medications and quite honestly, I wouldn’t have survived that long. They should be able to prescribe and be comfortable with that. Maybe additional training [options] for them. A certification or something. (P1023)

This participant highlighted the importance of practitioners being able to diagnose and treat the disease, which was a sentiment echoed by many other participants. Participants reported that the wait times to see a specialist were too long for a patient in crisis. Many advocated for training that would reach a wider set of practitioners to ease the barriers they faced in identification of PP and receiving treatment.

Policy level

At the policy level, participants advocated for government and healthcare institutional changes to better support those experiencing PP. Suggestions included paid family leave, insurance coverage for postpartum doulas, and expanded access to therapy. One participant shared, “Time and space to heal which means childcare support and time off from work” (P1322). Another participant noted that, “Improvements to public health that provide better support for families and reduce stress around issues of reproduction and healthcare would make a huge difference as well” (P1208). Lastly, one participant requested, “Better maternal Healthcare in general–postpartum doulas covered by insurance, paid time off, insurance having to cover all follow up care including therapy . . . ” (P1320). These requests speak to the vital policy changes which could be made at the company, state, or federal level to improve the pathway to recovery.

Society level

At the societal level, participants highlighted the need for reducing stigma and fostering open conversations about PP. Participants reported that societal norms or lack of awareness often hindered recovery, with one participant stating: “It would be helpful if the general public were aware, and the stigma of psychosis and mental illness were lessened, because that could encourage women to seek help without fear of judgement or ridicule” (P1086). Many participants emphasized the importance of sharing stories with more positive outcomes to challenge stigma and increase public awareness and knowledge of the illness.

Strengths and limitations

Strengths of this study include a large, geographically diverse sample from five countries, 33 US states, and Washington DC, as well as rigorous diagnostic confirmation using structured interviews in MGHP3. Additionally, the use of the SEM framework enhances the interpretive depth of findings, providing a structured lens for understanding the layered complexities of PP recovery. However, several limitations exist which should be considered when interpreting our findings. The sample lacked racial and socioeconomic diversity, with 87% identifying as white, 83% holding at least a bachelor’s degree, and 77% reporting private insurance during their perinatal period. The homogeneity of the sample impacts the generalizability to other racial/ethnic and socio-economic groups. Retrospective reporting may also introduce recall bias, although evidence suggests recall of traumatic experiences like PP is typically reliable (Doucet et al., 2012). Furthermore, the qualitative data were derived from short-answer, written survey responses, restricting the ability to probe deeper into participant experiences. The range of experiences and factors that individuals with lived experience of PP recovery were described in this content analysis, rather than a more interpretive thematic analysis that typically relies on in-depth interviews and richer narrative data. Finally, the 42% response rate to the supplemental survey invitation may have introduced response bias.

Future directions

Given the severity and complexity of PP, there is a need to expand awareness and multilevel supports for providers, patients, and families. Incorporating the voices of individuals with lived experience is crucial in designing patient-centered interventions. Future research could build on these reported findings and explore strategies to develop and implement comprehensive support services. Furthermore, the SEM was used as a framework to guide the analysis only; future research could consider using the SEM to guide qualitative inquiry during the development of an interview guide or survey, to intentionally explore constructs in the model. This study provides a preliminary mapping of patient-identified support needs which may contribute to a framework for future interventions. Findings underscore the importance of multi-level, theoretically grounded approaches to improving outcomes for individuals recovering from PP.