Mortality and clinical outcomes in peripartum cardiomyopathy: A systematic review and meta-analysis Authors: Resha Reya Ganthan, Francesco Rotatori Published: 2026 ABSTRACT BackgroundPeripartum cardiomyopathy (PPCM) is an important cause of maternal cardiovascular morbidity and mortality, characterized by the development of left ventricular systolic dysfunction toward the end of pregnancy or in the months following delivery. We aimed to evaluate all-cause mortality and clinical outcomes in women with PPCM.MethodsWe performed a systematic review and meta-analysis of studies evaluating outcomes in PPCM. PubMed, Embase, and Web of Science were searched from inception through April 2026. Studies reporting all-cause mortality and/or heart failure hospitalization were included. Pooled event rates were estimated using a random-effects model with logit transformation.ResultsA total of 60 studies comprising 44,922 patients were included. The pooled all-cause mortality rate was 7% (95% confidence interval [CI] 5–9%), with substantial heterogeneity (I2 = 96.4%). Reported mortality ranged from 0% to over 25%, reflecting marked variability across geographic regions and clinical settings. Heart failure hospitalization was reported in three studies; the pooled rate was 11% (95% CI 1–63%), although this estimate should be interpreted cautiously given the limited evidence base and substantial heterogeneity.ConclusionsPeripartum cardiomyopathy is associated with a meaningful risk of mortality, with substantial variability across populations. These findings highlight the need for improved recognition, risk stratification, and standardized longitudinal cardiovascular care in women with PPCM. Introduction Cardiovascular disease remains a leading cause of maternal morbidity and mortality worldwide, with an increasing proportion of adverse events occurring in the postpartum period [1], [2], [3]. While clinical attention has traditionally focused on complications during pregnancy and delivery, the weeks following childbirth—often referred to as the “fourth trimester”—represent a period of significant cardiovascular vulnerability characterized by dynamic hemodynamic and neurohormonal changes [4], [5], [6]. Peripartum cardiomyopathy (PPCM) is characterized by the development of left ventricular systolic dysfunction toward the end of pregnancy or in the months following delivery [7], [8], [9]. In addition to PPCM, heart failure in the postpartum setting may arise in association with hypertensive disorders of pregnancy, pre-existing cardiomyopathies, and cardiometabolic comorbidities, contributing to substantial clinical heterogeneity [10], [11], [12]. Outcomes in peripartum cardiomyopathy remain highly variable across studies and geographic regions. Reported mortality rates range from less than 5% in high-income settings to over 20% in resource-limited environments, reflecting differences in disease severity, access to care, and healthcare infrastructure [13], [14], [15], [16]. Furthermore, recovery of left ventricular function is inconsistent, with many patients experiencing persistent dysfunction and ongoing risk of adverse cardiovascular events [17], [18], [19]. Data on other clinically relevant outcomes, including heart failure hospitalization, are comparatively limited. Despite increasing recognition of peripartum cardiomyopathy, there remains a lack of comprehensive synthesis quantifying the burden of mortality and clinical outcomes across diverse populations. In this study, we performed a systematic review and meta-analysis to evaluate all-cause mortality and available data on clinical outcomes, including heart failure hospitalization, among patients with peripartum cardiomyopathy. This focus complements prior meta-analyses examining ventricular recovery and outcomes of subsequent pregnancy by specifically characterizing the magnitude and timing of mortality across contemporary and historical PPCM cohorts. Methods Search strategy and study selection This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A comprehensive literature search was performed across three electronic databases: PubMed, Embase, and Web of Science, from database inception through April 2026. The search strategy incorporated combinations of keywords and controlled vocabulary related to peripartum cardiomyopathy and heart failure. A total of 3541 records were identified (PubMed: 488; Web of Science: 590; Embase: 2463). After removal of duplicates, 2677 unique records were screened by title and abstract. Full-text assessment was performed for studies meeting initial eligibility criteria, with final study selection summarized in Fig. 1.Fig. 1PRISMA flow diagram of study selection. Records were identified through PubMed, Embase, and Web of Science, followed by removal of duplicates and screening of titles and abstracts. Full-text articles were assessed for eligibility, with exclusions based on predefined criteria.Fig. 1 PRISMA flow diagram of study selection. Records were identified through PubMed, Embase, and Web of Science, followed by removal of duplicates and screening of titles and abstracts. Full-text articles were assessed for eligibility, with exclusions based on predefined criteria. Eligibility criteria Studies were eligible for inclusion if they evaluated adult patients with peripartum cardiomyopathy and reported clinically relevant outcomes, including all-cause mortality and/or heart failure hospitalization. To be included in the quantitative synthesis, studies were required to provide extractable event-level data, including the number of outcome events and total sample size. Eligible study designs included observational cohort studies, registry-based analyses, and post hoc analyses of randomized controlled trials. Only studies involving human adult populations were considered. The primary outcome of interest was all-cause mortality, with HF hospitalization evaluated as a secondary clinical outcome where sufficient data were available. LVEF recovery and recurrence or relapse during subsequent pregnancy were not prespecified outcomes of the present review, as the objective was to quantify mortality burden and its variation across reported follow-up intervals rather than to reassess previously synthesized predictors of ventricular recovery or subsequent-pregnancy outcomes. Studies were excluded if they involved pediatric populations or non-human subjects, were case reports or small case series with fewer than 10 participants, or consisted of reviews, editorials, or conference abstracts without sufficient data for extraction. Studies that did not report relevant clinical outcomes or lacked sufficient data to derive outcome measures were also excluded. Data extraction and study characteristics Data were extracted using a standardized approach. Extracted variables included study identifier (author and year), sample size, and number of outcome events. The primary outcome was all-cause mortality, defined as death from any cause during the reported follow-up period. Where studies reported cause-specific mortality, only all-cause mortality was used for the primary quantitative synthesis. Cardiovascular and non-cardiovascular mortality were not pooled separately because cause-specific mortality was inconsistently reported across the included studies. The characteristics of included studies, including study identifier, geographic region, study design, sample size, population definition, and reported outcomes, are summarized in Table 1. Individual studies are identified by author and year and cross-referenced to their corresponding citations. A complete list of included studies with detailed study characteristics and outcome contributions is provided in Supplementary Table S1. Potentially overlapping cohorts were reviewed at the study level using study setting, data source, enrollment period, and population characteristics. Where substantial overlap was identified, only the most comprehensive or clinically relevant report was retained for the corresponding outcome. Distinct analyses with non-identical study populations or exposure definitions were retained.Table 1Characteristics of included studies. The table summarizes study design, geographic region, sample size, population definitions, and reported outcomes for studies evaluating peripartum cardiomyopathy.StudyRegionStudy designNPopulation definitionOutcomesAl Riyami et al., 2023OmanRetrospective cohort116PPCMMortalityBarasa et al., 2017SwedenRegistry-based cohort241HF in late pregnancy/postpartumMortalityBasic et al., 2022SwedenRegistry-based cohort187PPCM (≤5 months postpartum)MortalityBlauwet et al., 2016USAProspective cohort100PPCM (≤13 weeks postpartum)MortalityFarhan et al., 2021IraqProspective cohort64PPCM (≤6 months postpartum)MortalityHaghikia et al., 2013GermanyProspective cohort115PPCMMortalityHammami et al., 2024TunisiaRetrospective cohort27PPCMMortalityHayrapetyan et al., 2025ArmeniaMulticenter cohort98PPCMMortalityHilfiker-Kleiner et al., 2017EuropeRCT63PPCMMortalityKanakasabapathy et al., 2022IndiaProspective cohort32PPCMMortalityKaraye et al., 2020NigeriaMulticenter cohort244PPCM (postpartum)Mortality; HF hospitalizationKayikcioglu et al., 2024TurkeyRegistry-based cohort293PPCMMortalityLindley et al., 2017USARetrospective cohort39PPCMMortalityMajella et al., 2019IndiaProspective cohort14PPCMMortalityMoulig et al., 2019GermanyCohort66PPCMMortalityNepal et al., 2017NepalProspective cohort22PPCMMortalityNtusi et al., 2015South AfricaComparative cohort30PPCMMortalityPillarisetti et al., 2014USARetrospective cohort100PPCMMortalityMcNamara et al., 2015USAProspective cohort100PPCMMortalitySliwa et al., 2010GermanyRCT20PPCMMortalitySliwa et al., 2025GlobalProspective registry98PPCMMortalityJackson et al., 2024GlobalProspective registry535PPCMMortality; rehospitalizationMoulig et al., 2018GermanyComparative cohort70PPCMMortalityBernstein et al., 2001USARetrospective cohort23PPCMMortalityBrar et al., 2007USAPopulation-based cohort60PPCMMortalityBriasoulis et al., 2016USACohort47PPCMMortality; HF hospitalizationCénac et al., 2003NigerCohort50PPCMMortalityAshraf et al., 2023IndiaProspective cohort22PPCMMortalityDasgupta et al., 2025IndiaProspective cohort25PPCMMortality; HF hospitalizationDjordjevic et al., 2021EuropeRegistry cohort16Severe PPCM (LVAD)MortalityDrissa et al., 2019TunisiaRetrospective cohort20PPCMMortalityFarah et al., 2016USARetrospective cohort20PPCMMortalityHasan et al., 2010PakistanCohort32PPCMMortalityHorgan et al., 2013Northern EuropeRetrospective cohort12PPCMMortalityHyder et al., 2026USARetrospective cohort733PPCM with cardiogenic shockMortalityIsezuo et al., 2007NigeriaCohort65PPCMMortalityJhansi et al., 2024IndiaRetrospective cohort12PPCMMortalityKamdem et al., 2023CameroonRetrospective cohort29PPCMMortalityKayikcioglu et al., 2018TurkeyRegistry cohort301PPCMMortalityKitsou et al., 2026NorwayCase-control15PPCMMortalityKrishnamoorthy et al., 2016IndiaNational database4871PPCMMortalityLaw et al., 2025Hong KongMulticenter retrospective30PPCMMortalityLibhaber et al., 2015South AfricaCohort206PPCMMortalityLicciardi et al., 2025ItalyRegistry cohort28PPCMMortalityMace et al., 2025USARetrospective cohort256PPCMMortalityMumtaz et al., 2024IndiaCohort19PPCMMortalityNaibé et al., 2024Burkina FasoProspective cohort94PPCMMortalityOgah et al., 2025NigeriaCohort69PPCMMortalityOng et al., 2023MalaysiaRetrospective cohort29PPCMMortalityParcha et al., 2024USADatabase cohort856PPCMMortalityPattoli et al., 2024aUSANational database19,135PPCMMortalityPattoli et al., 2024bUSANational database13,540PPCMMortalityRoy et al., 2023IndiaProspective cohort20PPCMMortalitySa'idu et al., 2021NigeriaProspective registry406PPCMMortalitySharma et al., 2023IndiaCohort23PPCMMortalitySimeon et al., 2006NigeriaProspective cohort65PPCMMortalitySinkey et al., 2022USARetrospective cohort95PPCMMortalityVally et al., 2022FranceRetrospective cohort40PPCMMortalityWu et al., 2017TaiwanNational database925PPCMMortalityZahidi et al., 2023MoroccoRetrospective cohort88PPCMMortality Table 1Characteristics of included studies. The table summarizes study design, geographic region, sample size, population definitions, and reported outcomes for studies evaluating peripartum cardiomyopathy.StudyRegionStudy designNPopulation definitionOutcomesAl Riyami et al., 2023OmanRetrospective cohort116PPCMMortalityBarasa et al., 2017SwedenRegistry-based cohort241HF in late pregnancy/postpartumMortalityBasic et al., 2022SwedenRegistry-based cohort187PPCM (≤5 months postpartum)MortalityBlauwet et al., 2016USAProspective cohort100PPCM (≤13 weeks postpartum)MortalityFarhan et al., 2021IraqProspective cohort64PPCM (≤6 months postpartum)MortalityHaghikia et al., 2013GermanyProspective cohort115PPCMMortalityHammami et al., 2024TunisiaRetrospective cohort27PPCMMortalityHayrapetyan et al., 2025ArmeniaMulticenter cohort98PPCMMortalityHilfiker-Kleiner et al., 2017EuropeRCT63PPCMMortalityKanakasabapathy et al., 2022IndiaProspective cohort32PPCMMortalityKaraye et al., 2020NigeriaMulticenter cohort244PPCM (postpartum)Mortality; HF hospitalizationKayikcioglu et al., 2024TurkeyRegistry-based cohort293PPCMMortalityLindley et al., 2017USARetrospective cohort39PPCMMortalityMajella et al., 2019IndiaProspective cohort14PPCMMortalityMoulig et al., 2019GermanyCohort66PPCMMortalityNepal et al., 2017NepalProspective cohort22PPCMMortalityNtusi et al., 2015South AfricaComparative cohort30PPCMMortalityPillarisetti et al., 2014USARetrospective cohort100PPCMMortalityMcNamara et al., 2015USAProspective cohort100PPCMMortalitySliwa et al., 2010GermanyRCT20PPCMMortalitySliwa et al., 2025GlobalProspective registry98PPCMMortalityJackson et al., 2024GlobalProspective registry535PPCMMortality; rehospitalizationMoulig et al., 2018GermanyComparative cohort70PPCMMortalityBernstein et al., 2001USARetrospective cohort23PPCMMortalityBrar et al., 2007USAPopulation-based cohort60PPCMMortalityBriasoulis et al., 2016USACohort47PPCMMortality; HF hospitalizationCénac et al., 2003NigerCohort50PPCMMortalityAshraf et al., 2023IndiaProspective cohort22PPCMMortalityDasgupta et al., 2025IndiaProspective cohort25PPCMMortality; HF hospitalizationDjordjevic et al., 2021EuropeRegistry cohort16Severe PPCM (LVAD)MortalityDrissa et al., 2019TunisiaRetrospective cohort20PPCMMortalityFarah et al., 2016USARetrospective cohort20PPCMMortalityHasan et al., 2010PakistanCohort32PPCMMortalityHorgan et al., 2013Northern EuropeRetrospective cohort12PPCMMortalityHyder et al., 2026USARetrospective cohort733PPCM with cardiogenic shockMortalityIsezuo et al., 2007NigeriaCohort65PPCMMortalityJhansi et al., 2024IndiaRetrospective cohort12PPCMMortalityKamdem et al., 2023CameroonRetrospective cohort29PPCMMortalityKayikcioglu et al., 2018TurkeyRegistry cohort301PPCMMortalityKitsou et al., 2026NorwayCase-control15PPCMMortalityKrishnamoorthy et al., 2016IndiaNational database4871PPCMMortalityLaw et al., 2025Hong KongMulticenter retrospective30PPCMMortalityLibhaber et al., 2015South AfricaCohort206PPCMMortalityLicciardi et al., 2025ItalyRegistry cohort28PPCMMortalityMace et al., 2025USARetrospective cohort256PPCMMortalityMumtaz et al., 2024IndiaCohort19PPCMMortalityNaibé et al., 2024Burkina FasoProspective cohort94PPCMMortalityOgah et al., 2025NigeriaCohort69PPCMMortalityOng et al., 2023MalaysiaRetrospective cohort29PPCMMortalityParcha et al., 2024USADatabase cohort856PPCMMortalityPattoli et al., 2024aUSANational database19,135PPCMMortalityPattoli et al., 2024bUSANational database13,540PPCMMortalityRoy et al., 2023IndiaProspective cohort20PPCMMortalitySa'idu et al., 2021NigeriaProspective registry406PPCMMortalitySharma et al., 2023IndiaCohort23PPCMMortalitySimeon et al., 2006NigeriaProspective cohort65PPCMMortalitySinkey et al., 2022USARetrospective cohort95PPCMMortalityVally et al., 2022FranceRetrospective cohort40PPCMMortalityWu et al., 2017TaiwanNational database925PPCMMortalityZahidi et al., 2023MoroccoRetrospective cohort88PPCMMortality Characteristics of included studies. The table summarizes study design, geographic region, sample size, population definitions, and reported outcomes for studies evaluating peripartum cardiomyopathy. Quality assessment The methodological quality of included observational studies was assessed using the Newcastle–Ottawa Scale (NOS), which evaluates studies across three domains: selection, comparability, and outcome. Scores were assigned according to the standard NOS criteria, with a maximum of 9 points. Studies scoring ≥5 were considered of acceptable methodological quality for the primary synthesis, whereas studies scoring <5 were considered lower quality. Individual domain scores and total NOS scores are presented in Table 2.Table 2Quality assessment of included studies using the Newcastle–Ottawa Scale (NOS).StudySelectionComparabilityOutcomeTotalRiyami 20233126Barasa 20174239Basic 20224239Blauwet 20164239Farhan 20213126Haghikia 20134239Hammami 20243126Hayrapetyan 20253126Hilfiker-Kleiner 20174239Kanakasabapathy 20223126Karaye 20204239Kayikcioglu 20244239Lindley 20173126Majella 20193126Moulig 20193126Nepal 20173126Ntusi 20154239Pillarisetti 20143126Schelbert 20174239Sliwa 20104239Sliwa 20254239Van der Meer 20244239Moulig 20183126Bernstein 20013126Brar 20074239Briasoulis 20164239Cénac 20033126Cénac 19982024Dasgupta 20253126Djordjevic 20213126Drissa 20193126Farah 20163126Hasan 20103126Horgan 20133126Hyder 20264239Isezuo 20073126Jhansi 20243126Kamdem 20233126Kayikcioglu 20184239Kitsou 20263126Krishnamoorthy 20164239Law 20253126Lihhaber 20154239Licciardi 20254239Mace 20253126Mumtaz 20243126Naibé 20243126Ogah 20253126Ong 20233126Parcha 20244239Pattoli 20244239Roy 20233126Sa'idu 20214239Sharma 20233126Simeon 20063126Sinkey 20224239Vally 20223126Wu 20174239Zahidi 20233126Studies were evaluated across three domains: selection, comparability, and outcome, with higher scores indicating better methodological quality. Table 2Quality assessment of included studies using the Newcastle–Ottawa Scale (NOS).StudySelectionComparabilityOutcomeTotalRiyami 20233126Barasa 20174239Basic 20224239Blauwet 20164239Farhan 20213126Haghikia 20134239Hammami 20243126Hayrapetyan 20253126Hilfiker-Kleiner 20174239Kanakasabapathy 20223126Karaye 20204239Kayikcioglu 20244239Lindley 20173126Majella 20193126Moulig 20193126Nepal 20173126Ntusi 20154239Pillarisetti 20143126Schelbert 20174239Sliwa 20104239Sliwa 20254239Van der Meer 20244239Moulig 20183126Bernstein 20013126Brar 20074239Briasoulis 20164239Cénac 20033126Cénac 19982024Dasgupta 20253126Djordjevic 20213126Drissa 20193126Farah 20163126Hasan 20103126Horgan 20133126Hyder 20264239Isezuo 20073126Jhansi 20243126Kamdem 20233126Kayikcioglu 20184239Kitsou 20263126Krishnamoorthy 20164239Law 20253126Lihhaber 20154239Licciardi 20254239Mace 20253126Mumtaz 20243126Naibé 20243126Ogah 20253126Ong 20233126Parcha 20244239Pattoli 20244239Roy 20233126Sa'idu 20214239Sharma 20233126Simeon 20063126Sinkey 20224239Vally 20223126Wu 20174239Zahidi 20233126Studies were evaluated across three domains: selection, comparability, and outcome, with higher scores indicating better methodological quality. Quality assessment of included studies using the Newcastle–Ottawa Scale (NOS). Studies were evaluated across three domains: selection, comparability, and outcome, with higher scores indicating better methodological quality. One study [31] had an NOS score < 5 and was excluded from the revised primary quantitative synthesis. Statistical analysis Meta-analyses were conducted using R with the meta package. Pooled estimates of proportions were calculated using a random-effects model with logit transformation (PLOGIT). Between-study heterogeneity was assessed using the I2 statistic and Cochran's Q test. Random-effects confidence intervals were calculated using the Hartung–Knapp method. Sensitivity analyses were performed using a leave-one-out approach to evaluate the influence of individual studies on pooled estimates. Publication bias was assessed visually using funnel plots and statistically using Egger's regression test. Given that fewer than 10 studies contributed to the HF hospitalization analysis, formal assessment of publication bias for this outcome was not performed because funnel plot asymmetry tests have limited statistical power and are not considered reliable in meta-analyses with small numbers of studies. To evaluate the influence of outcome-assessment timing, a stratified analysis was performed according to mortality follow-up. Studies were categorized as reporting early mortality (≤6 months) or later mortality (>6 months) based on the reported mortality assessment time point. Studies without a clearly defined mortality follow-up interval were retained in the primary overall mortality analysis but excluded from the timing-stratified analysis. Differences between subgroups were assessed using a formal test for subgroup differences. Results Study selection and characteristics A total of 60 studies were included in the primary meta-analysis of mortality, comprising 44,922 patients and 1408 events. The included studies encompassed a wide range of geographic regions and clinical settings, with variability in study design, sample size, and duration of follow-up. Three studies were included in the secondary analysis evaluating heart failure (HF) hospitalization. Primary outcome: all-cause mortality The pooled all-cause mortality rate was 7.0% (95% CI 5.0%–9.0%) using a random-effects model (Fig. 2).Fig. 2Forest plot of pooled all-cause mortality in peripartum cardiomyopathy. Each square represents the estimate from an individual study, with horizontal lines indicating 95% confidence intervals. The diamond represents the pooled estimate derived using a random-effects model. Overall pooled all-cause mortality was 7% (95% CI 5–9%), with substantial heterogeneity across studies (I2 = 96.4%).Fig. 2 Forest plot of pooled all-cause mortality in peripartum cardiomyopathy. Each square represents the estimate from an individual study, with horizontal lines indicating 95% confidence intervals. The diamond represents the pooled estimate derived using a random-effects model. Overall pooled all-cause mortality was 7% (95% CI 5–9%), with substantial heterogeneity across studies (I2 = 96.4%). There was considerable variability in mortality estimates across studies, ranging from 0% to over 25%. Substantial heterogeneity was observed (I2 = 96.4%, p < 0.001), indicating significant differences in study populations and clinical characteristics. Mortality according to follow-up duration Among the 60 studies included in the primary mortality analysis, 43 had a clearly defined mortality assessment time point and were included in the timing-stratified analysis. Seventeen studies comprising 39,091 patients contributed to the early mortality (≤6 months) subgroup, with a pooled mortality rate of 3.4% (95% CI 1.7%–6.4%; I2 = 97.3%). Twenty-six studies comprising 3847 patients contributed to the later mortality (>6 months) subgroup, with a pooled mortality rate of 8.1% (95% CI 5.9%–10.9%; I2 = 77.1%). Mortality was significantly higher in studies reporting outcomes beyond 6 months compared with those reporting outcomes at or before 6 months (p for subgroup difference = 0.009). Seventeen studies without a clearly defined mortality assessment time point were excluded from this stratified analysis but remained included in the overall mortality synthesis (Fig. 3).Fig. 3Timing-stratified analysis of all-cause mortality in peripartum cardiomyopathy. Studies were stratified according to reported mortality assessment timing as early mortality (≤6 months) or later mortality (>6 months). Seventeen studies comprising 39,091 patients were included in the early mortality subgroup, with a pooled mortality rate of 3.4% (95% CI 1.7%–6.4%; I2 = 97.3%). Twenty-six studies comprising 3847 patients were included in the later mortality subgroup, with a pooled mortality rate of 8.1% (95% CI 5.9%–10.9%; I2 = 77.1%). Mortality was significantly higher in the >6-month subgroup (p for subgroup difference = 0.009). Studies without a clearly defined mortality assessment time point were excluded from this stratified analysis but retained in the overall mortality synthesis.Fig. 3 Timing-stratified analysis of all-cause mortality in peripartum cardiomyopathy. Studies were stratified according to reported mortality assessment timing as early mortality (≤6 months) or later mortality (>6 months). Seventeen studies comprising 39,091 patients were included in the early mortality subgroup, with a pooled mortality rate of 3.4% (95% CI 1.7%–6.4%; I2 = 97.3%). Twenty-six studies comprising 3847 patients were included in the later mortality subgroup, with a pooled mortality rate of 8.1% (95% CI 5.9%–10.9%; I2 = 77.1%). Mortality was significantly higher in the >6-month subgroup (p for subgroup difference = 0.009). Studies without a clearly defined mortality assessment time point were excluded from this stratified analysis but retained in the overall mortality synthesis. Secondary outcome: heart failure hospitalization A total of three studies comprising 316 patients contributed to the meta-analysis of HF hospitalization (Fig. 4). The pooled HF hospitalization rate was 11% (95% CI 1%–63%) using a random-effects model, with substantial heterogeneity (I2 = 90.7%, p < 0.001). Given the small number of contributing studies and wide confidence interval, this estimate should be interpreted cautiously.Fig. 4Forest plot of heart failure hospitalization in peripartum cardiomyopathy. Three studies comprising 316 patients contributed to the analysis. The pooled HF hospitalization rate was 11% (95% CI 1–63%), with substantial heterogeneity (I2 = 90.7%). Given the limited number of studies and wide confidence interval, these findings should be interpreted cautiously.Fig. 4 Forest plot of heart failure hospitalization in peripartum cardiomyopathy. Three studies comprising 316 patients contributed to the analysis. The pooled HF hospitalization rate was 11% (95% CI 1–63%), with substantial heterogeneity (I2 = 90.7%). Given the limited number of studies and wide confidence interval, these findings should be interpreted cautiously. Sensitivity analyses Leave-one-out sensitivity analysis demonstrated that exclusion of any individual study did not significantly alter the pooled mortality estimate, supporting the robustness of the primary findings (Fig. 5).Fig. 5Leave-one-out sensitivity analysis for pooled all-cause mortality in peripartum cardiomyopathy. Each iteration represents the pooled estimate after sequential exclusion of a single study. The pooled mortality estimate remained approximately 7% across all iterations, with 95% confidence intervals consistently overlapping the overall estimate, indicating that no individual study disproportionately influenced the primary result.Fig. 5 Leave-one-out sensitivity analysis for pooled all-cause mortality in peripartum cardiomyopathy. Each iteration represents the pooled estimate after sequential exclusion of a single study. The pooled mortality estimate remained approximately 7% across all iterations, with 95% confidence intervals consistently overlapping the overall estimate, indicating that no individual study disproportionately influenced the primary result. In an additional sensitivity analysis excluding two highly selected cohorts comprising patients with cardiogenic shock or requiring durable mechanical circulatory support, the pooled all-cause mortality estimate remained similar at 6.5% (95% CI 5.0%–8.4%; I2 = 96.1%; 58 studies, 44,173 patients), indicating that inclusion of these severe cohorts did not materially influence the primary estimate. Publication bias Funnel plot inspection suggested potential asymmetry (Fig. 6). Egger's regression test demonstrated evidence of small-study effects (t = 2.34, df = 58, p = 0.023). However, interpretation of this finding is limited by the substantial between-study heterogeneity.Fig. 6Funnel plot assessing potential publication bias for all-cause mortality in peripartum cardiomyopathy. Each point represents an individual study plotted against its standard error. Visual asymmetry was observed, and Egger's regression test demonstrated evidence of small-study effects (t = 2.34, df = 58, p = 0.023). Given the substantial between-study heterogeneity, this finding should be interpreted cautiously, as funnel plot asymmetry may reflect underlying clinical and methodological variability rather than publication bias alone.Fig. 6 Funnel plot assessing potential publication bias for all-cause mortality in peripartum cardiomyopathy. Each point represents an individual study plotted against its standard error. Visual asymmetry was observed, and Egger's regression test demonstrated evidence of small-study effects (t = 2.34, df = 58, p = 0.023). Given the substantial between-study heterogeneity, this finding should be interpreted cautiously, as funnel plot asymmetry may reflect underlying clinical and methodological variability rather than publication bias alone. Discussion Principal findings In this systematic review and meta-analysis of 60 studies comprising 44,922 patients, the pooled all-cause mortality rate in peripartum cardiomyopathy was 7.0%. Despite this overall estimate, there was substantial variability across studies, with mortality ranging from 0% to over 25%, and marked heterogeneity (I2 = 96.4%). In the secondary analysis, heart failure hospitalization was evaluated in a limited number of studies, and while a pooled estimate was derived, these findings should be interpreted cautiously. Comparison with existing literature Peripartum cardiomyopathy (PPCM) is a heterogeneous disorder with widely variable outcomes influenced by geographic region, access to care, and baseline disease severity [1], [2], [3], [4]. Prior studies have reported mortality rates ranging from less than 5% in high-income countries to over 20% in resource-limited settings [2], [5], [6], [7]. The pooled estimate observed in the present analysis is consistent with this spectrum and reflects the inclusion of diverse global populations. Differences in outcomes are likely multifactorial. Delayed diagnosis, limited access to guideline-directed medical therapy, and more advanced disease at presentation have been reported in low and middle income countries (LMICs) [5], [6], [8]. In contrast, earlier recognition and multidisciplinary care in high-income settings have been associated with improved outcomes [2], [3], [9]. Socioeconomic factors, healthcare infrastructure, and access to advanced therapies further contribute to disparities in outcomes [6], [8]. In addition, emerging data suggest that underlying pathophysiological mechanisms—including vascular dysfunction, inflammation, and oxidative stress—may influence disease severity and recovery, potentially contributing to observed variability across populations [22], [23], [24]. Clinical implications The findings of this study highlight that peripartum cardiomyopathy remains associated with a meaningful risk of mortality, even in contemporary cohorts [3], [9], [10]. The substantial variation in outcomes underscores the need for improved risk stratification and standardized management strategies. These findings support the integration of cardiovascular evaluation into routine postpartum care pathways. Recovery of left ventricular function is variable, with a significant proportion of patients experiencing persistent dysfunction [9], [10], [11]. Early identification of high-risk patients is therefore essential, particularly given the potential for adverse outcomes including arrhythmias, thromboembolism, and progressive heart failure [3], [10], [12]. The observed burden of HF hospitalization, although based on limited data, suggests ongoing morbidity in this population. Importantly, these findings support the concept of the postpartum period as a high-risk window for cardiovascular complications. The “fourth trimester” is characterized by dynamic hemodynamic and neurohormonal changes that may precipitate or exacerbate heart failure in susceptible individuals [13], [14], [25]. Despite this, structured postpartum cardiovascular follow-up remains inconsistent, and opportunities for early intervention may be missed [26], [27]. These results highlight the need for dedicated postpartum surveillance strategies and multidisciplinary care pathways extending beyond delivery [28], [29], [30]. Heterogeneity and study variability A key finding of this analysis is the high degree of heterogeneity across studies. This likely reflects differences in study design, patient populations, diagnostic criteria, and follow-up duration [15], [16]. The timing of outcome assessment varied substantially, with some studies reporting short-term outcomes and others evaluating longer-term mortality. Geographic variability is a major contributor. Studies conducted in sub-Saharan Africa and other LMICs frequently report higher mortality rates, consistent with prior literature [5], [6], [8]. In addition, variability in baseline left ventricular function, comorbidities, and access to therapies such as bromocriptine and advanced heart failure interventions may further contribute to outcome differences [2], [9], [17], [22]. The high degree of heterogeneity underscores the need for standardized definitions and reporting in peripartum cardiomyopathy research. Publication bias and small-study effects Funnel plot asymmetry and Egger's regression test suggested the presence of small-study effects. However, interpretation of these findings must be approached with caution given the substantial heterogeneity observed. In highly heterogeneous meta-analyses, funnel plot asymmetry may reflect underlying clinical variability rather than true publication bias [18], [19], [20], [23]. An additional source of heterogeneity is the broad temporal span of the included literature. PPCM management has evolved substantially over time, including greater use of contemporary guideline-directed heart failure therapy, improved access to multidisciplinary cardio-obstetric care, and broader availability of temporary and durable mechanical circulatory support. Accordingly, outcomes reported in earlier cohorts may not fully reflect those achievable in contemporary practice. Differences in treatment era and access to advanced heart failure therapies should therefore be considered when interpreting the pooled mortality estimates. Strengths and limitations This study has several strengths. It represents a large, comprehensive synthesis of over 44,000 patients across diverse clinical settings, providing a broad overview of outcomes in peripartum cardiomyopathy. The use of a random-effects model with logit transformation and robust sensitivity analyses enhances the reliability of the findings [20], [21], [22]. Several limitations should be acknowledged. First, the high heterogeneity limits the precision of pooled estimates. Second, the observational nature of included studies introduces potential bias and confounding [15], [16]. Third, variability in outcome definitions and follow-up duration may affect comparability across studies. Cause-specific mortality and associations between baseline or follow-up LVEF and mortality could not be quantitatively synthesized because these data were inconsistently defined and reported across the included studies. Future prospective PPCM studies should systematically report cardiovascular and non-cardiovascular causes of death and standardized measures of ventricular function to better characterize determinants of long-term mortality. Finally, the limited number of studies evaluating HF hospitalization restricts the strength of conclusions for this outcome. Conclusions In this systematic review and meta-analysis of PPCM, pooled all-cause mortality was approximately 7%, with substantially higher mortality reported beyond 6 months than during earlier follow-up. These findings highlight the importance of sustained longitudinal surveillance beyond the immediate peripartum period and demonstrate that outcome assessment timing is an important source of variability across PPCM studies. Although HF hospitalization data were limited, the available evidence further supports the need for continued cardiovascular follow-up after diagnosis. CRediT authorship contribution statement Resha Reya Ganthan: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Francesco Rotatori: Writing – review & editing, Supervision, Methodology. Consent Obtained. Consent for publication Not applicable. Ethics approval and consent to participate Not applicable. This study is a systematic review and meta-analysis of previously published data and did not involve the collection of new individual-level human or animal data. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.