A machine learning approach to identifying pregnant women's risk for persistent post-traumatic stress following childbirth

A scoping review of effective health practices for the treatment of birth trauma

BACKGROUND

There is currently no consensus on the most effective health practices to manage or reduce the effects of birth trauma (BT) and childbirth-related posttraumatic stress disorder (CB-PTSD).

AIM

The aim was to map the current literature on effective health practices for BT/CB-PTSD, identify key elements (the what, when and how) important for effective health practices, and highlight gaps in maternity care.

METHODS

A systematic search was conducted across key nursing, allied, and medical databases (MEDLINE, Scopus, PubMed) for key terms related to (1) birth trauma and (2) intervention. Only peer-reviewed, English-language papers published since 2000 were included to ensure the relevance and timeliness of the findings. Following PRISMA-ScR guidelines, 6,347 articles were identified through databases/registers and citation searching. After removing 1,342 duplicates, 5,005 were screened by title and abstract. A further 4,544 were excluded, leaving 461 for full-text screening. Afterf excluding another 433, 28 papers met inclusion for this review.

FINDINGS

The first session delivered early (within the first 72 h of birth) by a clinician (midwife/psychologist/counsellor) significantly reduced BT/CB-PTSD in the short-term. Both trauma-focused and non-trauma-focused were supported at this stage, provided they were structured. If intervention is delayed (weeks to months post-birth), a trauma-focused, multi-session approach is recommended.

DISCUSSION

Early, structured interventions should be considered routine care for women with BT/CB-PTSD, with more intensive, structured, trauma-focused approach for persistent symptoms. The potential role of digital mental health tools is promising, particularly for women in low-resource settings, but requires further research to evaluate feasibility, acceptability, and sustainability.

Precise multi-factor immediate implant placement decision models based on machine learning

This study aims to explore the effect of implant apex design, osteotomy preparation, intraosseous depth and bone quality on immediate implant placement insertion torque and establish a more sophisticated decision model with multi-factor analysis based on machine learning for improving the success rate of immediate implant placement. Six implant replicas of each of the three implant systems with different implant apex design were placed in polyurethane foam block with different densities(soft, medium and hard) via two osteotomy preparation protocols (normal preparation and undersized preparation) at different implant intraosseous depths (3 mm, 5 mm and 7 mm). The insertion torque for each implant was recorded and subsequently analyzed using one-way and four-way ANOVA. Prediction models of insertion torque were then constructed using multiple linear regression (MLR) and decision tree regression (DTR) analyses based on multi-factors. These machine learning models were evaluated and compared for their predictive accuracy and performance. The influencing factors of immedate implant placement insertion torque are ranked as follows: bone quality, intraosseous depth, osteotomy preparation protocol, and implant apex design. Both two machine learning preoperative prediction models (MLR and DTR) showed high accuracy in insertion torque prediction, with the latter's R2 reaching as high as 0.951. This research is of significant reference value for optimizing clinical decision-making, improving the success rate of immediate implant placement, and enhancing the efficiency of doctor-patient communication. In addition, this study further refined the evaluation framework for implant performance, rendering it more comprehensive and standardized.

A systematic review of machine learning findings in PTSD and their relationships with theoretical models

In recent years, the application of machine learning (ML) techniques in research on the prediction of post-traumatic stress disorder (PTSD) has increased. However, concerns regarding the clinical relevance and generalizability of ML findings hamper their implementation by clinicians and researchers. Here in this systematic review we examined (1) the extent to which pre-, peri- and post-traumatic risk factors identified using ML approaches coincide with the theoretical understanding of the disorder; (2) whether new insights were gained through ML techniques; and (3) whether ML findings, combined with previous research, enable an integrative model of PTSD risk encompassing both predictor categories and their theoretical relevance. We reviewed ML studies on PTSD risk factors in PubMed, Web of Science and Scopus. Studies were included if they specified when predictors and PTSD symptoms were collected in temporal relation to the traumatic event. A total of 30 studies with 12,908 participants (mean age 36.5 years) were included. After extracting the 15 most important predictors from all studies, we categorized them into pre-, peri- and post-trauma exposure predictors and examined their associations with established theoretical models of PTSD. Many studies exhibited a risk of bias, assessed using the prediction model risk of bias assessment tool (PROBAST). However, we found overlaps in identified predictors across studies, a concordance between data-driven results and theory-driven research, and underexplored predictors identified through ML. We propose an integrative model of PTSD risk that incorporates both data-driven and theory-driven findings and discuss future directions. We emphasize the importance of standards on how to apply and report ML approaches for mental health.

Exploring and Predicting HIV Preexposure Prophylaxis Adherence Patterns Among Men Who Have Sex With Men: Randomized Controlled Longitudinal Study of an mHealth Intervention in Western China

Background

Preexposure prophylaxis (PrEP) is an effective strategy to reduce the risk of HIV infection. However, the efficacy of PrEP is highly dependent on adherence. Meanwhile, adherence changes over time, making it difficult to manage effectively.

Objective

Our study aimed to explore and predict the patterns of change in PrEP adherence among men who have sex with men (MSM) and evaluate the impact of the WeChat-based reminder intervention on adherence, thus providing more information for PrEP implementation strategies.

Methods

From November 2019 to June 2023, in a randomized controlled longitudinal study of the PrEP demonstration project in Western China (Chongqing, Sichuan, and Xinjiang) based on a mobile health (mHealth) reminder intervention, participants were randomly divided into reminder and no-reminder groups, with those in the reminder group receiving daily reminders based on the WeChat app. Participants were followed up and self-reported their medication adherence every 12 weeks for a total of 5 follow-up visits. We used the growth mixture model (GMM) to explore potential categories and longitudinal trajectories of adherence among MSM, and patterns of change in PrEP adherence were predicted and evaluated based on the decision tree.

Results

A total of 446 MSM were included in the analysis. The GMM identified 3 trajectories of adherence: intermediate adherence group (n=34, 7.62%), low adherence ascending group (n=126, 28.25%), and high adherence decline group (n=286, 64.13%). We included 8 variables that were significant in the univariate analysis in the decision tree prediction model. We found 4 factors and 8 prediction rules, and the results showed that HIV knowledge score, education attainment, mHealth intervention, and HIV testing were key nodes in the patterns of change in adherence. After 10-fold cross-validation, the final prediction model had an accuracy of 75%, and the classification accuracy of low and intermediate adherence was 78.12%.

Conclusions

The WeChat-based reminder intervention was beneficial for adherence. A short set of questions and prediction rules, which can be applied in future large-scale validation studies, aimed at developing and validating a short adherence assessment tool and implementing it in PrEP practices among MSM.

A Machine Learning-Based Prediction Model for the Probability of Fall Risk Among Chinese Community-Dwelling Older Adults

Fall is a common adverse event among older adults. This study aimed to identify essential fall factors and develop a machine learning-based prediction model to predict the fall risk category among community-dwelling older adults, leading to earlier intervention and better outcomes. Three prediction models (logistic regression, random forest, and naive Bayes) were constructed and evaluated. A total of 459 people were involved, including 156 participants (34.0%) with high fall risk. Seven independent predictors (frail status, age, smoking, heart attack, cerebrovascular disease, arthritis, and osteoporosis) were selected to develop the models. Among the three machine learning models, the logistic regression model had the best model fit, with the highest area under the curve (0.856) and accuracy (0.797) and sensitivity (0.735) in the test set. The logistic regression model had excellent discrimination, calibration, and clinical decision-making ability, which could aid in accurately identifying the high-risk groups and taking early intervention with the model.

Persistent symptoms of maternal post-traumatic stress following childbirth across the first months postpartum: Associations with perturbations in maternal behavior and infant avoidance of social gaze toward mother

Recent literature identifies childbirth as a potentially traumatic event, following which mothers may develop symptoms of Post-Traumatic-Stress-Following-Childbirth (PTS-FC). The current study examines whether stable symptoms of PTS-FC during the early postpartum period may impose risk for perturbations in maternal behavior and infant social-engagement with mother, controlling for comorbid postpartum internalizing symptoms. Mother-infant dyads (N = 192) were recruited from the general population, during the third trimester of pregnancy. 49.5% of the mothers were primipara, and 48.4% of the infants were girls. Maternal PTS-FC was assessed at 3-day, 1-month and 4-month postpartum, via self-report and clinician-administered interview. Latent Profile Analysis generated two profiles of symptomology: "Stable-High-PTS-FC" (17.0%), and "Stable-Low-PTS-FC" (83%). Membership in the "Stable-High-PTS-FC" profile associated with perturbed maternal sensitivity, which was in turn significantly associated with infant avoidance of social gaze toward mother (Indirect effect β = -0.15). Results suggest the need for early screening and inform the planning of early preventive interventions.

Identifying women with postdelivery posttraumatic stress disorder using natural language processing of personal childbirth narratives

BACKGROUND

Maternal mental disorders are considered a leading complication of childbirth and a common contributor to maternal death. In addition to undermining maternal welfare, untreated postpartum psychopathology can result in child emotional and physical neglect and associated significant pediatric health costs. Some women may experience traumatic childbirth and develop posttraumatic stress disorder symptoms after delivery (childbirth-related posttraumatic stress disorder). Although women are routinely screened for postpartum depression in the United States, there is no recommended protocol to inform the identification of women who are likely to experience childbirth-related posttraumatic stress disorder. Advancements in computational methods of free text have shown promise in informing the diagnosis of psychiatric conditions. Although the language in narratives of stressful events has been associated with posttrauma outcomes, whether the narratives of childbirth processed via machine learning can be useful for childbirth-related posttraumatic stress disorder screening is unknown.

OBJECTIVE

This study aimed to examine the use of written narrative accounts of personal childbirth experiences for the identification of women with childbirth-related posttraumatic stress disorder. To this end, we developed a model based on natural language processing and machine learning algorithms to identify childbirth-related posttraumatic stress disorder via the classification of birth narratives.

STUDY DESIGN

Overall, 1127 eligible postpartum women who enrolled in a study survey during the COVID-19 pandemic provided short written childbirth narrative accounts in which they were instructed to focus on the most distressing aspects of their childbirth experience. They also completed a posttraumatic stress disorder symptom screen to determine childbirth-related posttraumatic stress disorder. After the exclusion criteria were applied, data from 995 participants were analyzed. A machine learning-based Sentence-Transformers natural language processing model was used to represent narratives as vectors that served as inputs for a neural network machine learning model developed in this study to identify participants with childbirth-related posttraumatic stress disorder.

RESULTS

The machine learning model derived from natural language processing of childbirth narratives achieved good performance (area under the curve, 0.75; F1 score, 0.76; sensitivity, 0.8; specificity, 0.70). Moreover, women with childbirth-related posttraumatic stress disorder generated longer narratives (t test results: t=2.30; p=.02) and used more negative emotional expressions (Wilcoxon test: sadness: p=8.90e-04; W=31,017; anger: p=1.32e-02; W=35,005.50) and death-related words (Wilcoxon test: p=3.48e-05; W=34,538) in describing their childbirth experience than those with no childbirth-related posttraumatic stress disorder.

CONCLUSION

This study provided proof of concept that personal childbirth narrative accounts generated in the early postpartum period and analyzed via advanced computational methods can detect with relatively high accuracy women who are likely to endorse childbirth-related posttraumatic stress disorder and those at low risk. This suggests that birth narratives could be promising for informing low-cost, noninvasive tools for maternal mental health screening, and more research that used machine learning to predict early signs of maternal psychiatric morbidity is warranted.

Identifying Women with Post-Delivery Posttraumatic Stress Disorder using Natural Language Processing of Personal Childbirth Narratives

Background

Maternal mental disorders are considered a leading complication of childbirth and a common contributor to maternal death. In addition to undermining maternal welfare, untreated postpartum psychopathology can result in child emotional and physical neglect, and associated significant pediatric health costs. Some women may experience a traumatic childbirth and develop posttraumatic stress disorder (PTSD) symptoms following delivery (CB-PTSD). Although women are routinely screened for postpartum depression in the U.S., there is no recommended protocol to inform the identification of women who are likely to experience CB-PTSD. Advancements in computational methods of free text has shown promise in informing diagnosis of psychiatric conditions. Although the language in narratives of stressful events has been associated with post-trauma outcomes, whether the narratives of childbirth processed via machine learning can be useful for CB-PTSD screening is unknown.

Objective

This study examined the utility of written narrative accounts of personal childbirth experience for the identification of women with provisional CB-PTSD. To this end, we developed a model based on natural language processing (NLP) and machine learning (ML) algorithms to identify CB-PTSD via classification of birth narratives.

Study Design

A total of 1,127 eligible postpartum women who enrolled in a study survey during the COVID-19 era provided short written childbirth narrative accounts in which they were instructed to focus on the most distressing aspects of their childbirth experience. They also completed a PTSD symptom screen to determine provisional CB-PTSD. After exclusion criteria were applied, data from 995 participants was analyzed. An ML-based Sentence-Transformer NLP model was used to represent narratives as vectors that served as inputs for a neural network ML model developed in this study to identify participants with provisional CB-PTSD.

Results

The ML model derived from NLP of childbirth narratives achieved good performance: AUC 0.75, F1-score 0.76, sensitivity 0.8, and specificity 0.70. Moreover, women with provisional CB-PTSD generated longer narratives (t-test results: t=2 . 30, p=0 . 02 ) and used more negative emotional expressions (Wilcoxon test: 'sadness': p=8 . 90e- ⁰⁴ , W=31,017 ; 'anger': p=1 . 32e- ⁰² , W=35,005 . 50 ) and death-related words (Wilcoxon test: p=3 . 48e- ⁰⁵ , W=34,538 ) in describing their childbirth experience than those with no CB-PTSD.

Conclusions

This study provides proof of concept that personal childbirth narrative accounts generated in the early postpartum period and analyzed via advanced computational methods can detect with relatively high accuracy women who are likely to endorse CB-PTSD and those at low risk. This suggests that birth narratives could be promising for informing low-cost, non-invasive tools for maternal mental health screening, and more research that utilizes ML to predict early signs of maternal psychiatric morbidity is warranted.

Detection of the most influential variables for preventing postpartum urinary incontinence using machine learning techniques

Background

Postpartum urinary incontinence is a fairly widespread health problem in today’s society among women who have given birth. Recent studies analysing the different variables that may be related to Postpartum urinary incontinence have brought to light some variables that may be related to Postpartum urinary incontinence in order to try to prevent it. However, no studies have been found that analyse some of the intrinsic and extrinsic variables of patients during pregnancy that could give rise to this pathology.

Objective

The objective of this study is to assess the most influential variables in Postpartum urinary incontinence by means of machine learning techniques, starting from a group of intrinsic variables, another group of extrinsic variables and a mixed group that combines both types.

Methods

Information was collected on 93 patients, pregnant women who gave birth. Experiments were conducted using different machine learning classification techniques combined with oversampling techniques to predict four variables: urinary incontinence, urinary incontinence frequency, urinary incontinence intensity and stress urinary incontinence.

Results

The results showed that the most accurate predictive models were those trained with extrinsic variables, obtaining accuracy values of 70% for urinary incontinence, 77% for urinary incontinence frequency, 71% for urinary incontinence intensity and 93% for stress urinary incontinence.

Conclusions

This research has shown that extrinsic variables are more important than intrinsic variables in predicting problems related to postpartum urinary incontinence. Therefore, although not conclusive, it opens a line of research that could confirm that the prevention of Postpartum urinary incontinence could be achieved by following healthy habits in pregnant women.