Understanding Symptom Self-Monitoring Needs Among Postpartum Black Patients: Qualitative Interview Study.

BACKGROUND: Pregnancy-related death is on the rise in the United States, and there are significant disparities in outcomes for Black patients. Most solutions that address pregnancy-related death are hospital based, which rely on patients recognizing symptoms and seeking care from a health system, an area where many Black patients have reported experiencing bias. There is a need for patient-centered solutions that support and encourage postpartum people to seek care for severe symptoms. OBJECTIVE: We aimed to determine the design needs for a mobile health (mHealth) patient-reported outcomes and decision-support system to assist Black patients in assessing when to seek medical care for severe postpartum symptoms. These findings may also support different perinatal populations and minoritized groups in other clinical settings. METHODS: We conducted semistructured interviews with 36 participants-15 (42%) obstetric health professionals, 10 (28%) mental health professionals, and 11 (31%) postpartum Black patients. The interview questions included the following: current practices for symptom monitoring, barriers to and facilitators of effective monitoring, and design requirements for an mHealth system that supports monitoring for severe symptoms. Interviews were audio recorded and transcribed. We analyzed transcripts using directed content analysis and the constant comparative process. We adopted a thematic analysis approach, eliciting themes deductively using conceptual frameworks from health behavior and human information processing, while also allowing new themes to inductively arise from the data. Our team involved multiple coders to promote reliability through a consensus process. RESULTS: Our findings revealed considerations related to relevant symptom inputs for postpartum support, the drivers that may affect symptom processing, and the design needs for symptom self-monitoring and patient decision-support interventions. First, participants viewed both somatic and psychological symptom inputs as important to capture. Second, self-perception; previous experience; sociocultural, financial, environmental, and health systems-level factors were all perceived to impact how patients processed, made decisions about, and acted upon their symptoms. Third, participants provided recommendations for system design that involved allowing for user control and freedom. They also stressed the importance of careful wording of decision-support messages, such that messages that recommend them to seek care convey urgency but do not provoke anxiety. Alternatively, messages that recommend they may not need care should make the patient feel heard and reassured. CONCLUSIONS: Future solutions for postpartum symptom monitoring should include both somatic and psychological symptoms, which may require combining existing measures to elicit symptoms in a nuanced manner. Solutions should allow for varied, safe interactions to suit individual needs. While mHealth or other apps may not be able to address all the social or financial needs of a person, they may at least provide information, so that patients can easily access other supportive resources.

Preparing for the bedside-optimizing a postpartum depression risk prediction model for clinical implementation in a health system.

OBJECTIVE: We developed and externally validated a machine-learning model to predict postpartum depression (PPD) using data from electronic health records (EHRs). Effort is under way to implement the PPD prediction model within the EHR system for clinical decision support. We describe the pre-implementation evaluation process that considered model performance, fairness, and clinical appropriateness. MATERIALS AND METHODS: We used EHR data from an academic medical center (AMC) and a clinical research network database from 2014 to 2020 to evaluate the predictive performance and net benefit of the PPD risk model. We used area under the curve and sensitivity as predictive performance and conducted a decision curve analysis. In assessing model fairness, we employed metrics such as disparate impact, equal opportunity, and predictive parity with the White race being the privileged value. The model was also reviewed by multidisciplinary experts for clinical appropriateness. Lastly, we debiased the model by comparing 5 different debiasing approaches of fairness through blindness and reweighing. RESULTS: We determined the classification threshold through a performance evaluation that prioritized sensitivity and decision curve analysis. The baseline PPD model exhibited some unfairness in the AMC data but had a fair performance in the clinical research network data. We revised the model by fairness through blindness, a debiasing approach that yielded the best overall performance and fairness, while considering clinical appropriateness suggested by the expert reviewers. DISCUSSION AND CONCLUSION: The findings emphasize the need for a thorough evaluation of intervention-specific models, considering predictive performance, fairness, and appropriateness before clinical implementation.

Visualizing machine learning-based predictions of postpartum depression risk for lay audiences.

OBJECTIVES: To determine if different formats for conveying machine learning (ML)-derived postpartum depression risks impact patient classification of recommended actions (primary outcome) and intention to seek care, perceived risk, trust, and preferences (secondary outcomes). MATERIALS AND METHODS: We recruited English-speaking females of childbearing age (18-45 years) using an online survey platform. We created 2 exposure variables (presentation format and risk severity), each with 4 levels, manipulated within-subject. Presentation formats consisted of text only, numeric only, gradient number line, and segmented number line. For each format viewed, participants answered questions regarding each outcome. RESULTS: Five hundred four participants (mean age 31 years) completed the survey. For the risk classification question, performance was high (93%) with no significant differences between presentation formats. There were main effects of risk level (all P < .001) such that participants perceived higher risk, were more likely to agree to treatment, and more trusting in their obstetrics team as the risk level increased, but we found inconsistencies in which presentation format corresponded to the highest perceived risk, trust, or behavioral intention. The gradient number line was the most preferred format (43%). DISCUSSION AND CONCLUSION: All formats resulted high accuracy related to the classification outcome (primary), but there were nuanced differences in risk perceptions, behavioral intentions, and trust. Investigators should choose health data visualizations based on the primary goal they want lay audiences to accomplish with the ML risk score.

Women's perspectives on the use of artificial intelligence (AI)-based technologies in mental healthcare.

This study aimed to evaluate women's attitudes towards artificial intelligence (AI)-based technologies used in mental health care. We conducted a cross-sectional, online survey of U.S. adults reporting female sex at birth focused on bioethical considerations for AI-based technologies in mental healthcare, stratifying by previous pregnancy. Survey respondents (n = 258) were open to AI-based technologies in mental healthcare but concerned about medical harm and inappropriate data sharing. They held clinicians, developers, healthcare systems, and the government responsible for harm. Most reported it was "very important" for them to understand AI output. More previously pregnant respondents reported being told AI played a small role in mental healthcare was "very important" versus those not previously pregnant (P = .03). We conclude that protections against harm, transparency around data use, preservation of the patient-clinician relationship, and patient comprehension of AI predictions may facilitate trust in AI-based technologies for mental healthcare among women.

Recentering responsible and explainable artificial intelligence research on patients: implications in perinatal psychiatry.

In the setting of underdiagnosed and undertreated perinatal depression (PD), Artificial intelligence (AI) solutions are poised to help predict and treat PD. In the near future, perinatal patients may interact with AI during clinical decision-making, in their patient portals, or through AI-powered chatbots delivering psychotherapy. The increase in potential AI applications has led to discussions regarding responsible AI and explainable AI (XAI). Current discussions of RAI, however, are limited in their consideration of the patient as an active participant with AI. Therefore, we propose a patient-centered, rather than a patient-adjacent, approach to RAI and XAI, that identifies autonomy, beneficence, justice, trust, privacy, and transparency as core concepts to uphold for health professionals and patients. We present empirical evidence that these principles are strongly valued by patients. We further suggest possible design solutions that uphold these principles and acknowledge the pressing need for further research about practical applications to uphold these principles.

Identifying urban built environment factors in pregnancy care and maternal mental health outcomes.

BACKGROUNDS: Risk factors related to the built environment have been associated with women's mental health and preventive care. This study sought to identify built environment factors that are associated with variations in prenatal care and subsequent pregnancy-related outcomes in an urban setting. METHODS: In a retrospective observational study, we characterized the types and frequency of prenatal care events that are associated with the various built environment factors of the patients' residing neighborhoods. In comparison to women living in higher-quality built environments, we hypothesize that women who reside in lower-quality built environments experience different patterns of clinical events that may increase the risk for adverse outcomes. Using machine learning, we performed pattern detection to characterize the variability in prenatal care concerning encounter types, clinical problems, and medication prescriptions. Structural equation modeling was used to test the associations among built environment, prenatal care variation, and pregnancy outcome. The main outcome is postpartum depression (PPD) diagnosis within 1 year following childbirth. The exposures were the quality of the built environment in the patients' residing neighborhoods. Electronic health records (EHR) data of pregnant women (n = 8,949) who had live delivery at an urban academic medical center from 2015 to 2017 were included in the study. RESULTS: We discovered prenatal care patterns that were summarized into three common types. Women who experienced the prenatal care pattern with the highest rates of PPD were more likely to reside in neighborhoods with homogeneous land use, lower walkability, lower air pollutant concentration, and lower retail floor ratios after adjusting for age, neighborhood average education level, marital status, and income inequality. CONCLUSIONS: In an urban setting, multi-purpose and walkable communities were found to be associated with a lower risk of PPD. Findings may inform urban design policies and provide awareness for care providers on the association of patients' residing neighborhoods and healthy pregnancy.

Development and validation of a machine learning algorithm for predicting the risk of postpartum depression among pregnant women.

OBJECTIVE: There is a scarcity in tools to predict postpartum depression (PPD). We propose a machine learning framework for PPD risk prediction using data extracted from electronic health records (EHRs). METHODS: Two EHR datasets containing data on 15,197 women from 2015 to 2018 at a single site, and 53,972 women from 2004 to 2017 at multiple sites were used as development and validation sets, respectively, to construct the PPD risk prediction model. The primary outcome was a diagnosis of PPD within 1 year following childbirth. A framework of data extraction, processing, and machine learning was implemented to select a minimal list of features from the EHR datasets to ensure model performance and to enable future point-of-care risk prediction. RESULTS: The best-performing model uses from clinical features related to mental health history, medical comorbidity, obstetric complications, medication prescription orders, and patient demographic characteristics. The model performances as measured by area under the receiver operating characteristic curve (AUC) are 0.937 (95% CI 0.912 - 0.962) and 0.886 (95% CI 0.879-0.893) in the development and validation datasets, respectively. The model performances were consistent when tested using data ending at multiple time periods during pregnancy and at childbirth. LIMITATIONS: The prevalence of PPD in the study data represented a treatment prevalence and is likely lower than the illness prevalence. CONCLUSIONS: EHRs and machine learning offer the ability to identify women at risk for PPD early in their pregnancy. This may facilitate scalable and timely prevention and intervention, reducing negative outcomes and the associated burden.

Lithium Use and Non-use for Pregnant and Postpartum Women with Bipolar Disorder.

PURPOSE OF REVIEW: Despite being recognized as a first-line treatment for bipolar disorder, there is still inconsistent use of lithium in perinatal populations. This article will review data regarding lithium use during the peripartum and provide management recommendations for general psychiatric clinicians. RECENT FINDINGS: In contrast to prior data, recent studies indicate that lithium use in pregnancy is associated with either no increased malformations risk or a small increase in risk for cardiac malformations including Ebstein's anomaly. Limited data also show no significant effect on obstetric or neurodevelopmental outcomes. Data regarding infant lithium exposure via breastmilk remains limited. Lithium is currently under-prescribed and is an important treatment for women with bipolar disorder in pregnancy and the postpartum. Clinicians must weigh the risk of lithium treatment versus the risk of withholding or changing lithium treatment when managing bipolar disorder in this population.

GABAA dysregulation as an explanatory model for late-onset postpartum depression associated with weaning and resumption of menstruation.

It is well established that a subgroup of women are particularly vulnerable to affective dysregulation during times of hormonal fluctuation. One underrecognized reproductive transition may be late-onset postpartum depression (PPD) in the context of weaning from breastfeeding and the resumption of menstruation. The goal of this review is to propose a biologically plausible mechanism for affective dysregulation during these transitions. The relationship between affective symptoms and neurohormonal changes associated with weaning will be investigated through a hypothesis-driven review of relevant literature. Neurosteroids, like allopregnanolone (ALLO), are widely recognized for augmenting GABAergic inhibition and having a powerful anxiolytic effect (Belelli D and Lambert JL, Nature Reviews Neuroscience 6:565-575, 2005). However, when ALLO is administered after prolonged withdrawal, there may be a paradoxical anxiogenic effect (Smith et al., Psychopharmacology 186:323-333, 2006; Shen et al., Nat Neurosci 10:469-477, 2007). Weaning from breastfeeding is a physiologic example of fluctuating levels of ALLO after prolonged withdrawal. We propose that the complex hormonal milieu during weaning and resumption of menstruation may modify GABAA receptors such that ALLO may contribute to rather than ameliorate depressive symptoms in vulnerable individuals. The proposed model provides an initial step for understanding the mechanisms by which the changing hormonal environment during weaning and resumption of menstruation may contribute to an increased risk of depression in a subgroup of women who are hormonally sensitive. Future research investigating this model would be valuable both to identify women at increased risk for developing mood symptoms late in postpartum and to inform treatment for this and related reproductive depressive disorders.

Protection against hormone-mediated mood symptoms.

We present the case of a woman with bipolar I disorder with severe premenstrual mood instability, confusion, and psychosis resembling the clinical features of postpartum psychosis when estrogen levels are expected to be low, and hypomania when estrogen levels are expected to be elevated. While depressive symptoms across the menstrual cycle have been extensively documented in the literature, there is little information regarding manic and hypomanic symptoms. In addition, we describe the successful treatment of her menstrual-cycle related symptoms. Approaches to the management of menstrual psychosis have not been systematically studied, and clinical guidelines do not exist. Clinical experiences such as the one reported here, in which the clinical formulation of the patient was consistent with known neuroendocrine phenomena and in which the treatment approach was successful, are crucial to developing promising approaches that can be tested in controlled trials.

Selective serotonin reuptake inhibitors for depression in pregnancy.

Perinatal depression is associated with a high risk of morbidity and mortality and may have long-term consequences on child development. The US Preventive Services Task Force has recently recognized the importance of identifying and treating women with depression in the perinatal period. However, screening and accessing appropriate treatment come with logistical challenges. In many areas, there may not be sufficient access to psychiatric care, and, until these resources develop, the burden may inadvertently fall on obstetricians. As a result, understanding the risks of perinatal depression in comparison with the risks of treatment is important. Many studies of selective serotonin reuptake inhibitors in pregnancy fail to control for underlying depressive illness, which can lead to misinterpretation of selective serotonin reuptake inhibitor risk by clinicians. This review discusses the risks and benefits of selective serotonin reuptake inhibitor treatment in pregnancy within the context of perinatal depression. Whereas selective serotonin reuptake inhibitors may be associated with certain risks, the absolute risks are low and may be outweighed by the risks of untreated depression for many women and their offspring.

A simple, efficient tool for assessment of mice after unilateral cortex injury.

A refined battery of neurological tests, SNAP (Simple Neuroassessment of Asymmetric Impairment), was developed and validated to efficiently assess neurological deficits induced in a mouse model of traumatic brain injury. Four to 7-month old mice were subjected to unilateral controlled cortical impact or sham injury (craniectomy only). Several behavioral tests (SNAP, beam walk, foot fault, and water maze) were used to assess functional deficits. SNAP was unique among these in that it required no expensive equipment and was performed in less than 5 min per mouse. SNAP demonstrated a high level of sensitivity and specificity as determined by receiver-operator characteristics curve analysis. Interrater reliability was good, as determined by Cohen's Kappa method and by comparing the sensitivity and specificity across various raters. SNAP detected deficits in proprioception, visual fields, and motor strength in brain-injured mice at 3 days, and was sensitive enough to detect magnitude and recovery of injury. The contribution of individual battery components changed as a function of time after injury, however, each was important to the overall SNAP score. SNAP provided a sensitive, reliable, time-efficient and cost-effective means of assessing neurological deficits in mice after unilateral brain injury.