Preventing posttraumatic stress disorder following childbirth: a systematic review and meta-analysis

Abstract

Objective: Women can develop posttraumatic stress disorder in response to experienced or perceived traumatic, often medically complicated, childbirth; the prevalence of these events remains high in the United States. Currently, no recommended treatment exists in routine care to prevent or mitigate maternal childbirth-related posttraumatic stress disorder. We conducted a systematic review and meta-analysis of clinical trials that evaluated any therapy to prevent or treat childbirth-related posttraumatic stress disorder.

Data sources: PsycInfo, PsycArticles, PubMed (MEDLINE), ClinicalTrials.gov, CINAHL, ProQuest, Sociological Abstracts, Google Scholar, Embase, Web of Science, ScienceDirect, Scopus, and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched for eligible trials published through September 2023.

Study eligibility criteria: Trials were included if they were interventional, if they evaluated any therapy for childbirth-related posttraumatic stress disorder for the indication of symptoms or before posttraumatic stress disorder onset, and if they were written in English.

Methods: Independent coders extracted the sample characteristics and intervention information of the eligible studies and evaluated the trials using the Downs and Black's quality checklist and Cochrane's method for risk of bias evaluation. Meta-analysis was conducted to evaluate pooled effect sizes of secondary and tertiary prevention trials.

Results: A total of 41 studies (32 randomized controlled trials, 9 nonrandomized trials) were reviewed. They evaluated brief psychological therapies including debriefing, trauma-focused therapies (including cognitive behavioral therapy and expressive writing), memory consolidation and reconsolidation blockage, mother-infant-focused therapies, and educational interventions. The trials targeted secondary preventions aimed at buffering childbirth-related posttraumatic stress disorder usually after traumatic childbirth (n=24), tertiary preventions among women with probable childbirth-related posttraumatic stress disorder (n=14), and primary prevention during pregnancy (n=3). A meta-analysis of the combined randomized secondary preventions showed moderate effects in reducing childbirth-related posttraumatic stress disorder symptoms when compared with usual treatment (standardized mean difference, -0.67; 95% confidence interval, -0.92 to -0.42). Single-session therapy within 96 hours of birth was helpful (standardized mean difference, -0.55). Brief, structured, trauma-focused therapies and semi-structured, midwife-led, dialogue-based psychological counseling showed the largest effects (standardized mean difference, -0.95 and -0.91, respectively). Other treatment approaches (eg, the Tetris game, mindfulness, mother-infant-focused treatment) warrant more research. Tertiary preventions produced smaller effects than secondary prevention but are potentially clinically meaningful (standardized mean difference, -0.37; -0.60 to -0.14). Antepartum educational approaches may help, but insufficient empirical evidence exists.

Conclusion: Brief trauma-focused and non-trauma-focused psychological therapies delivered early in the period following traumatic childbirth offer a critical and feasible opportunity to buffer the symptoms of childbirth-related posttraumatic stress disorder. Future research that integrates diagnostic and biological measures can inform treatment use and the mechanisms at work.

Keywords: Cesarean delivery; childbirth trauma; childbirth-related posttraumatic stress disorder (CB-PTSD); delivery; maternal morbidity; obstetrical complications; obstetrics; postpartum period; posttraumatic stress disorder (PTSD); psychological intervention.

Figure 1.. PRISMA flow diagram for CB-PTSD intervention articles selection

PRISMA flow diagram detailing the source selection process of both randomized and non-randomized clinical studies targeting childbirth-related post-traumatic stress disorder (CB-PTSD) in at-risk and universal samples.

Figure 2.. Treatment approaches for childbirth-related PTSD prevention

Primary, secondary, and tertiary preventive treatment approaches for CB-PTSD tested in the reviewed trials. Grey boxes indicate treatment modality, purple boxes highlight the most efficacious interventions, and white boxes indicate specific type of interventions. Trauma-focused expressive writing (TF-EW) for secondary prevention was tested in universal samples.

Figure 3.. Risk of bias for randomized trials for CB-PTSD prevention

Risk of bias assessment by intervention modality for all randomized trials targeting CB-PTSD prevention or treatment of CB-PTSD using the Cochrane RoB 2 tool.

Figure 4.. Risk of bias for non-randomized trials for CB-PTSD prevention

Risk of bias assessment by intervention modality for all non-randomized trials targeting CB-PTSD prevention or treatment of CB-PTSD using the Cochrane ROBINS-I tool.

Figure 5.. Meta-analysis of secondary prevention trials for CB-PTSD

Forest plot depicting the pooled effect estimates of CB-PTSD symptoms in the meta-analysis of 23 comparisons (n=4999). Each square represents the study-specific effect size, with the size of the square corresponding to the weight of the study in the analysis. Horizontal lines denote the 95% confidence interval (CI) for each study, and the diamond represents the overall pooled effect estimate with its 95% CI. The vertical line at the null value (mean difference of 0) indicates no effect. Studies favoring the intervention are to the left of the line, while those favoring the control are to the right. Heterogeneity among studies is assessed by the I² statistic. For the Deforges 2023 study, Mean and SD were generated from symptom counts (Figure 2 from that paper) using the PlotDigitizer platform (https://plotdigitizer.com).

Figure 6.. Subgroup analysis of secondary prevention trials for CB-PTSD by treatment modality

Forest plot illustrating subgroup analysis for treatment modality within the meta-analysis. The squares represent the effect estimates for each subgroup. Horizontal lines indicate the 95% confidence interval (CI) for each subgroup. The diamonds represent the pooled effect estimates for each subgroup, and their width reflects the corresponding 95% CI. The vertical line at the null value indicates no effect. Subgroup-specific effects are compared (test for subgroup differences). Heterogeneity within subgroups is assessed by the I² statistic, providing insights into the consistency of effects across different subpopulations. Overall heterogeneity is also provided.

Figure 7.. Subgroup analysis of secondary prevention trials for CB-PTSD by number of sessions

Forest plot illustrating subgroup analysis for number of sessions (> 1 session) within the meta-analysis. The squares represent the effect estimates for each subgroup. Horizontal lines indicate the 95% confidence interval (CI) for each subgroup. The diamonds represent the pooled effect estimates for each subgroup, and their width reflects the corresponding 95% CI. The vertical line at the null value indicates no effect. Subgroup-specific effects are compared (test for subgroup differences). Heterogeneity within subgroups is assessed by the I² statistic, providing insights into the consistency of effects across different subpopulations. Overall heterogeneity is also provided.

Figure 8.. Meta-analysis of tertiary prevention trials for CB-PTSD

A forest plot is utilized to display the aggregated effect estimates of CB-PTSD symptoms in a meta-analysis comprising 9 comparisons (n=813). Each square corresponds to the effect size specific to each study, and its size reflects the study’s weight in the analysis. The 95% confidence interval (CI) for each study is represented by horizontal lines, and the diamond illustrates the overall pooled effect estimate alongside its 95% CI. The vertical line at the null value (mean difference of 0) signifies no effect. Studies favoring the intervention are positioned to the left of the line, while those favoring the control are on the right. The I² statistic is employed to assess heterogeneity among the studies.

Figure 9.. Subgroup analysis of tertiary prevention trials for CB-PTSD by treatment modality

A forest plot is presented to depict the subgroup analysis based on treatment modalities within the meta-analysis. Each square corresponds to the effect estimate for a specific subgroup, with horizontal lines indicating the 95% confidence interval (CI) for each subgroup. Diamonds represent the pooled effect estimates for each subgroup, and their width indicates the corresponding 95% CI. The vertical line at the null value signifies no effect. The comparison of subgroup-specific effects involves testing for subgroup differences. Heterogeneity within subgroups is evaluated using the I² statistic, offering insights into the consistency of effects across diverse subpopulations. Additionally, the overall heterogeneity is reported.