Background Post-exposure prophylaxis (PEP) and dog vaccination continue to be the widely adopted vaccination strategy for rabies control in high-burden countries like India. To determine the public health and economic impacts of childhood pre-exposure prophylaxis (PrEP) relative to existing strategies, we carried out a modelling analysis in the state of Kerala, India. Methods We adapted a probabilistic decision-tree model over a 10-year time horizon (2026–2035). This timeframe is sufficient for dog vaccination to interrupt rabies transmission and for protection to accrue from PrEP delivered through routine childhood immunisation. The model simulates interactions between dog vaccination, PrEP, and PEP to estimate health provider costs, quality-adjusted life years (QALYs) gained, and net health benefit across a range of willingness-to-pay thresholds. Findings Assuming current practice of four-dose intradermal PEP and sub-optimal dog vaccination, we projected 431 rabies deaths (95% Prediction interval [PI] 353–523) and a cost of US$50.3 million over ten years, with dog vaccination responsible for almost 70% of deaths averted by this strategy. Routine childhood PrEP increased costs by US$10.5–36.5 million (INR 1.0–3.4 billion), depending on the regimen used, and cumulatively prevented only ten additional deaths beyond those averted by current practice. Incorporating PrEP was not optimal (probability ≤12%) across India's willingness-to-pay threshold. The intensified owned-dog vaccination strategy achieved the highest benefit at India's mean willingness-to-pay (739,420 QALYs; 28% probability of optimality), while adopting WHO's recommended abridged PEP regimen was the only cost-saving strategy (US$6.3 million saved; equivalent mortality to status quo). Comprehensive dog vaccination coverage of 70% combined with abridged PEP regimens became optimal at India's upper willingness-to-pay threshold, reducing deaths by 83% compared to current practice. Interpretation In high-PEP-access settings, PrEP is not cost-effective and provides negligible benefit. To maximise lives saved, PrEP should only be considered once gains from strengthening dog vaccination and PEP delivery have been exhausted. Funding Wellcome Trust; Indian Council of Medical Research - Department of Health Research (ICMR-DHR) International Fellowship for Senior Biomedical Scientists.
Background There has long been an interest in the association between family planning (contraception and abortion) and women’s empowerment, with mixed evidence as to an association and most studies relying on cross sectional data. Within this body of work, less focus has been on how family planning impacts women’s empowerment The aim of this study is to understand the impact of abortion denial resulting in having an unwanted birth on women’s empowerment, measured by education, mobility and decision-making. Methods In the Nepal Turnaway Study, we followed women seeking abortions from 2019 to 2020 in Nepal and assessed whether receiving versus being denied an abortion was negatively associated with women’s empowerment over five years (longitudinal cohort). We used linear and logistic mixed effects regression models to assess the effects of obtaining an abortion compared to carrying an unwanted pregnancy to term on education, mobility outside the home and household decision-making power. Findings We followed 1742 for 5 years and find that women who did not give birth were more likely to be in school (Risk Difference (RD) = 0.2 starting 18 months post-abortion seeking) and have greater decision-making power over five years (Risk Difference (RD) = −0.2 starting 12 months post-abortion seeking) and both of these continued for the remainder of the time period. We find no evidence that giving birth affects a woman’s freedom to leave her home for shopping or connections to family and friends. Interpretation These findings suggest that being able to decide the timing and circumstances of childbearing has wide-ranging effects on women’s empowerment (as measured by schooling and decision-making). Funding This study was supported by the National Institute of Health David and Lucile Packard Foundation.
Background Access to abortion is central to people’s ability to act on their fertility preferences, yet little research has examined whether the ability to avoid having an unwanted birth shapes subsequent fertility outcomes or enables people to align future childbearing with their preferences. This study examines whether and how access to a desired abortion is associated with fertility preferences and reproductive outcomes over the subsequent years. Methods We analyzed data from the Nepal Turnaway Study, a five-year prospective cohort study of women seeking abortion, comparing those who obtained an abortion with those who were denied and subsequently gave birth. We examined fertility preferences over time using mixed-effects multinomial logistic regression, time to subsequent pregnancy and birth using flexible parametric survival models, and cumulative births using mixed-effects Poisson regression. Finally, we assessed whether subsequent births were undesired prior to conception using mixed-effects logistic regression. Findings Among 1742 participants seeking abortion, 1467 had an abortion or miscarried (the no-birth-to-index-pregnancy group) and 275 were denied and gave birth (the gave-birth-to-index-pregnancy group). Initially (at year one through year four), the no-birth-to-index-pregnancy group was more open to pregnancy than the gave-birth-to-index-pregnancy group; by year five, their fertility preferences converged. Four years after return to fecundability, the cumulative incidence of a subsequent pregnancy was 41.8% in the no-birth-to-index-pregnancy group and 18.3% in the gave-birth-to-index-pregnancy group (difference: 23.5 percentage points; 95% CI: 17.3–29.7), and of a subsequent birth was 30.8% and 10.0%, respectively (difference: 20.8 percentage points; 95% CI: 15.1–26.6). The predicted mean total number of births, including the index birth, per woman by year five was 1.04 for the gave-birth-to-index-pregnancy group and 0.37 for the no-birth-to-index-pregnancy group (difference: 0.67; 95% CI: 0.56–0.79). A smaller proportion of post-index births were undesired (before conception) in the no-birth-to-index-pregnancy group compared with the gave-birth-to-index-pregnancy group (35.1% vs. 84.5%; difference: 49.4 percentage points; 95% CI: 33.8–65.0; p = 0.0001). Interpretation Access to abortion shapes subsequent fertility preferences and outcomes and can lead to greater alignment between future reproductive preferences and outcomes. Funding This study was supported by the National Institutes of Health and the David and Lucile Packard Foundation.
Background Women may have adverse physical health following abortion denial and childbirth, particularly where maternal morbidity is high. This study aims to investigate mobility, self-rated physical health, and chronic pain associated with unwanted pregnancy among women seeking facility-based abortion services in Nepal. Methods Women seeking abortion at 22 facilities across Nepal in 2019–2020 completed surveys every six months for five years. This longitudinal cohort study assessed limited mobility, self-rated poor health, and chronic pain using weighted linear and logistic mixed effects regression, including among those seeking abortion for health reasons. We report time-specific marginal predicted probabilities for those denied (who gave birth and did not give birth) and those who obtained abortion. Findings Of the sample (n = 1742), 51% obtained care (Abortions), 15% gave birth (Turnaway-Births), and 33% were denied but did not give birth (Turnaway-NoBirths). Risk of any limited mobility was 15 percentage points higher (95% CI: 7%, 22%, p = 0.0001) for Turnaway-Births (predicted probability: 47%) versus Abortions (32%), and 2.5 days longer on average (95% CI: 1.42, 3.59, p < 0.0001; 5.3 versus 2.8 days); differences were larger among those seeking abortion for health reasons. Self-rated poor health, pelvic pain, and head pain did not differ for Turnaway-Births versus Abortions overall. Among the subsample, Turnaway-Births had poorer self-rated health at 18 months (0.09, [0.01, 0.16], p = 0.026), more back pain at 24–36 months (maximum risk difference: 0.17 [0.04, 0.31], p = 0.012), and less pelvic pain at 6 months (−0.08, [−0.015, −0.01], p = 0.036) compared to the Abortions. Interpretation Compared to the Abortions, Turnaway-Births experienced more limited mobility, and, among those who sought abortion for health reasons, poorer self-rated health and more chronic pain. Support for women seeking abortion, especially for health reasons, in obtaining timely care is needed to limit adverse physical health. Funding This study was supported by the National Institutes of Health and the David and Lucile Packard Foundation.
Background The consequences of abortion denial for women's lives are increasingly recognized, but little is known about the well-being of children born as a result of denied abortion. We compared early childhood outcomes between children born due to abortion denial (index children) and children born from a subsequent pregnancy after their mother received an abortion (subsequent children). Methods Longitudinal data from the five-year prospective cohort Nepal Turnaway Study (2019–2024) include index children and subsequent children. We assessed nutritional status, health, development, and household conditions through maternal report and direct anthropometric measurement. We employed mixed effects models with propensity score weighting to balance differences in maternal characteristics at time of abortion-seeking. Findings Differences between index (n = 274) and subsequent (n = 421) children are to the detriment of index children. Disadvantage starts before birth when their mothers reported more limited diets in terms of green leafy vegetables and sufficient quantity of food. Index children are more likely than subsequent children to be born at home OR = 1.91 (95% CI: 1.01–3.61), less than 24 months from an older sibling OR = 21.26 (95% CI: 4.39–102.85) and at a birthweight below 2500g OR = 1.92 (95% CI: 1.17–3.16). In their first years of life, index children were more likely than subsequent children to live in households with inadequate income (42.3% vs 31.8% p = 0.009), experience acute malnutrition (25.3% vs 17.4% p = 0.011); and die (0.7% vs 0.2% p = 0.038). Interpretation Being born after abortion denial is associated with economic and nutritional disadvantage in early childhood, underscoring the intergenerational consequences of restricting access to abortion. Funding The U.S. National Institute of Health and the David and Lucile Packard Foundation.
Background Economic growth in low- and middle-income countries is usually associated with reductions in mortality; however, it also correlates with increased alcohol use. This rise contributes to mortality and other disease burden, potentially hindering further economic growth. We quantified these associations for the Association of Southeast Asian Nations (ASEAN) countries for the years 2000–2023. Methods Structural equation modelling (SEM) was used to analyse time-series data for nine ASEAN countries with full data on economic output (Gross Domestic Product [GDP] and Gross National Income per capita at purchasing power parity [PPP]), disease burden (mortality and disability-adjusted life years [DALYs]), and alcohol exposure indicators for adults aged 15 to 65. All data were obtained from publicly available international agency sources. Findings SEM results indicate a positive association between alcohol use indicators and disease burden, which was, in turn, negatively associated with economic output. In 2021, for adults aged 15–65, alcohol-attributable mortality amounted to 5% of all-cause mortality in the region with three countries exceeding 10%; proportions for DALYs were similar. In the main model, the point estimates of potential impact on GDP PPP per capita were as high as 4%. The results remained highly consistent, irrespective of the indicators used. Interpretation Alcohol consumption in ASEAN countries over the study period hindered the ability to reap the full benefits of economic growth. Funding Centre for Alcohol Studies, Prince of Songkla University, Thailand, CAS-A1-67-002; WHO Western Pacific Region (2024/1505643-0; PO 203515367); Canadian Institutes of Health Research (FRN 181677, and FRN 203773).
Background Leprosy remains a significant public health issue in endemic regions, including areas in Indonesia. Transmission of Mycobacterium leprae (M. leprae) continues as clear from the almost stable annual number of new leprosy cases amongst which around 5% are children. To interrupt transmission of M. leprae, it is essential to identify and treat infection sources. This study aimed to assess anti-M. leprae phenolic glycolipid-I (PGL-I) IgM seroprevalence among school-aged children in rural and urban Indonesia, as a proxy for recent transmission. Additionally, associations between seropositivity and gender, socioeconomic status (SES), and helminth infection were investigated. Methods We conducted a cross-sectional serosurvey among 637 children (ages 6–15) from schools in urban and rural areas of Sulawesi, Java, and Sumba, Indonesia. The PGL-I QURapid, a field-friendly and quantitative lateral flow assay, was used to detect anti-PGL-I IgM in fingerstick blood and plasma samples. Socioeconomic and health data, including helminth infection status and body-mass index z-scores (z-BMI), were also collected. Findings The overall seroprevalence of anti-PGL-I IgM in children was 12.2% with significantly higher seroprevalence in rural (Southwest Sumba: 31.0%, Pangkajene: 11.2%) compared to urban areas (North Jakarta: 3.8%, Makassar city: 3.1%). Moreover, seroprevalence amongst children from low-SES schools was significantly higher (13.7%) than among those from high-SES schools (4.1%, P = 0.007). The presence of helminth infections were associated with significantly higher anti-PGL-I IgM levels (P < 0.0001), with 22.4% of helminth-infected children testing seropositive compared to 8.8% of non-infected children (P < 0.001). Interpretation Our findings demonstrate a high anti-PGL-I IgM seroprevalence among rural, low-SES, and helminth-infected children in Indonesia, indicating elevated levels of M. leprae infection within these groups which again represents increased transmission rates in the population. These results underscore the need for integrated disease control strategies addressing both leprosy and helminthiasis, particularly in rural and economically disadvantaged areas. The PGL-I QURapid demonstrated to be an effective tool in serosurveys, able to aid the identification of areas with high M. leprae transmission rates (such as Sumba) and thereby support the implementation of targeted interventions in endemic regions. Funding This study was funded by a grant from the Q.M. Gastmann-Wichers Foundation (AG) and an internal grant from Atma Jaya Catholic University of Indonesia (MMMK).
National AI health strategies are proliferating, yet development methodologies remain largely undocumented, limiting collective learning. India's Strategy for AI in Healthcare (SAHI) was developed through a three-phase participatory process spanning situational analysis, regional consultations across four geographic zones, and structured synthesis engaging over 300 stakeholders. Documenting that process architecture for examination and adaptation by other countries is the methodological contribution of this paper. Consultation surfaced five system-level constraints recurring across emerging AI health ecosystems and drove a fundamental restructuring from prescriptive implementation planning to an enabling framework. A dual framework emerged separating technical prerequisites from governance and implementation conditions, with equity embedded as an operational requirement throughout. Rights-based regulation, market-led innovation, and state-centric control have dominated global AI governance debate. SAHI reflects a fourth approach, anchoring governance in shared digital public infrastructure rather than rules alone, and offers a distinct reference point for any country developing a health-specific AI strategy.
Background:India aims to eliminate measles and rubella by 2026. In this study we analysed surveillance data from 2016 to 2025 from a facility-based surveillance of 16 sentinel sites across India to understand the burden and trend of congenital rubella syndrome (CRS). Methods:CRS surveillance included eligible infants between 0 and 11 months of age with suspected CRS, defined as the presence of congenital heart defects, hearing impairment, ocular abnormalities or maternal history of rash during pregnancy. All enrolled infants underwent comprehensive clinical evaluation, including cardiac, ophthalmologic, and hearing assessments. Serum samples were tested for rubella-specific IgM and IgG antibodies. The final analysis summarised clinical characteristics and trend of laboratory-confirmed CRS infants. Findings:Between November 2016 and May 2025, 8953 infants with suspected CRS were enrolled across 16 sentinel sites, among them 583 (6.5%) patients were classified as having laboratory-confirmed CRS. The proportion of laboratory-confirmed patients declined from 24.9% in 2017 to 2.5% in 2025. Among laboratory-confirmed CRS patients, 81.3% had structural heart defects, 26.8% had hearing impairment, and 24.2% had any of the group A ocular sign (cataract, glaucoma or pigmentary retinopathy). Of the 583 infants with laboratory-confirmed CRS, 541 were tested for rubella virus by RT-PCR, of whom 56 (10.4%) were positive, with no positive detections after 2021. Follow-up information was available for 452 laboratory-confirmed CRS patients, of whom 126 died, with a cumulative probability of death by one year of 32.2% (95% CI 26.7-38.4). Interpretation:The observed decline in the proportion of laboratory-confirmed CRS among suspected CRS infants in this facility-based surveillance is consistent with declining rubella virus transmission in India, although changes in surveillance sensitivity over time could not be excluded. Sustained high-quality surveillance is essential to monitor progress towards CRS elimination in the country. Funding:CRS surveillance was funded by the Ministry of Health and Family Welfare, Govt of India, through Gavi, the Vaccine Alliance (Grant number: 00101970).
India is committed to the World Health Organization (WHO) and the International Agency for the Prevention of Blindness “2030 In Sight” recommendations to improve eye health and visual impairment. Public health experts in India were invited to discuss and recommend strategies to reduce blindness by 2030 and consolidate the gains related to quality eye care by 2047. The experts reviewed the World Report on Vision, WHO’s 2030 In Sight strategy, and the status of eye care in India, drawing on published literature and policy documents. They proposed 10 interlinked strategic pillars: universal coverage, human resources, financial sustainability, equity, partnerships, health information systems, quality care, elderly and marginalised eyecare, technology adoption, and monitoring of key indices. Using the Likert scale the experts arrived at consensus on the following recommendations: task shifting, measuring quality eyecare indicators, integrating health services, leveraging technology, expanding cashless services, and improving referral eyecare system. The recommendations also include target-oriented spending, centralised management of available resources, and interministerial commitment. Funding Hyderabad Eye Research Foundation, Hyderabad, India.
IgA nephropathy (IgAN) is a major cause of chronic kidney disease in South Asia, particularly among young adults. Although recent therapeutic advances have transformed IgAN management globally, access to these treatments in South Asia remains limited. Most patients continue to rely on supportive care and conventional immunosuppressives, which have limited efficacy and significant adverse effects. Multiple barriers hinder implementation of newer therapies, including shortage of healthcare capacity and specialist nephrology care leading to delayed diagnosis, high treatment costs, limited resources, lack of universal health coverage, regulatory challenges, and underrepresentation of South Asian populations in clinical trials. This viewpoint discusses the current landscape of IgAN management in South Asia and outlines strategies to improve equitable access to novel therapies, including strengthening healthcare infrastructure and workforce, expanding screening and diagnostic capacity, implementing pricing and reimbursement reforms, facilitating clinical trials and local evidence generation, improving disease awareness in the community as well as healthcare professionals, and enhancing stakeholder advocacy and policy engagement.
Kyasanur Forest Disease (KFD) is a tick-borne zoonotic infection endemic to certain parts of India, with expanding geographic distribution and recurrent focal outbreaks among forest-dependent populations in Western Ghats. The discontinuation of the formalin-inactivated KFD vaccine in 2022 has created a critical gap in prevention. New vaccine candidates are under development, but conventional Phase 3 efficacy trials are difficult to implement because of low incidence of disease, and clustered and unpredictable nature of outbreaks. KFD vaccine evaluation may require alternate regulatory approaches. An initial step would involve studies to identify candidate correlates of protection in laboratory confirmed KFD patients. These immune markers could then be evaluated in bonnet macaques challenge models to assess their association with protection against infection and clinical disease and define protective immune thresholds or composite immunological endpoints. A phase 3 immunobridging study could subsequently assess seroresponse rates against predefined immune thresholds and provide supportive evidence for vaccine efficacy and regulatory decision-making. Post-licensure studies can assess real world vaccine effectiveness and help refine vaccination strategies. Vaccine deployment should prioritise high-risk populations and align with seasonal transmission patterns of KFD. This framework may provide a feasible pathway for KFD vaccine deployment in endemic regions.
Background:The optimal duration of systemic antibiotic prophylaxis (SAP) for Gustilo-Anderson type IIIA and IIIB open tibial fractures remains uncertain. We compared 2-day versus 3-day prophylaxis in patients treated at a level I trauma centre. Methods:The Prophylactic Antibiotic Course in Open Tibial Fractures (PROACT) trial was a pragmatic, single-centre, single-blinded randomised controlled trial conducted in India. Adults (≥18 years) with Gustilo-Anderson type IIIA or IIIB open tibial fractures were randomly assigned (1:1) to receive 2 days or 3 days of intravenous cefuroxime and amikacin; metronidazole was added for farm-related contamination. Outcome assessors were masked, and all patients received standardised orthoplastic care. The primary outcome was 90-day surgical-site infection (SSI). Analyses were by intention to treat. The trial was prospectively registered with the Clinical Trials Registry of India (CTRI/2020/06/026083) and is closed to recruitment. Findings:Between July 1, 2020, and Dec 31, 2022, 241 patients were assessed for eligibility, and 204 were randomly assigned to 2-day (n = 104) or 3-day (n = 100) prophylaxis; all were included in the primary analysis. SSI within 90 days occurred in 15 of 104 patients (14%) and 18 of 100 patients (18%), respectively (risk difference -3.6%, 95% CI -14.0 to 7.0; p = 0.60). Time-to-event analysis showed no statistically significant difference between groups (hazard ratio 1.22, 95% CI 0.60-2.46). Unplanned reoperation within 12 months occurred in 24 of 104 patients (23%) and 19 of 100 patients (19%). One patient in each group underwent secondary amputation due to flap failure. No treatment-related deaths were reported. Interpretation:Among adults with severe open tibial fractures managed with contemporary orthoplastic care, 2-day compared with 3-day SAP was not associated with a statistically significant difference in SSI. However, clinically meaningful benefit or harm cannot be excluded because of the limited sample size and imprecise estimates. Larger multicentre studies are warranted to confirm these findings. Funding:This study received no external funding.
Background Antimicrobial stewardship (AMS) implementation is highly context dependent. This mixed-methods study aimed to assess contextual factors associated with AMS implementation at national, facility, and healthcare-provider levels in four countries. Methods We used a sequential mixed-methods approach, with a combination of quantitative surveys, qualitative interviews and focus group discussions (FGDs) using the World Health Organisation's National AMS assessment tools at national, healthcare-facility, and provider levels in Indonesia, Nepal, Thailand, and Viet Nam between 2022 and 2023. Facility-level implementation was evaluated using the U.S. Centers for Disease Control and Prevention's Global Antibiotic Stewardship Evaluation Tool (G-ASET) in 16 hospitals across these countries. We included stakeholders and healthcare personnel involved in AMS implementation at the national and hospital levels. Findings All four countries had national AMS policies, and two incorporated financing mechanisms into national action plans. Main areas for further improvement were the development of stewardship guidelines for primary care and community settings; stronger linkages across related sectors and programmes; expanded professional training; strengthened regulatory enforcement; systematic performance tracking; and better data use for feedback and decision-making. Among the 16 hospitals, the mean G-ASET score was 241/335 (72%), highest in tracking, monitoring and reporting [56/75 (75%)] and lowest in training and education [17/25 (66%)], and the scores varied by hospital type and size. Although all hospitals had initiated AMS activities, implementation was often partial, with gaps in data-informed treatment guidelines, staff training, and resources. Interpretation National guidelines on treatment and stewardship initiatives are crucial and trainings for healthcare personnel need to be strengthened for facility-level AMS implementation. Enhancing inter-programme coordination and support for small-size hospitals and primary care settings in Asia will be essential to advance AMS implementation. Funding Supported by the U.S. Centers for Disease Control and Prevention.
Background:India contributed to 17.3% of the global stillbirth burden in 2019 and anaemia, an important risk factor for stillbirth, affects nearly half of pregnant women in the country. To understand the relationship between maternal anaemia and stillbirth, we analysed a diverse multicohort dataset from the Stillbirth Pooled Indian Cohort Consortium (ICMR-SPIC) from India. Methods:This was a secondary analysis of the ICMR-SPIC dataset, which includes individual-level data from eight observational studies and two randomised controlled trials conducted across ten states in India. Most cohorts recruited participants between 2010 and 2023, except for the Pune Maternal Nutrition Study (PMNS), which recruited participants between 1994 and 1996. A total of 214,709 singleton pregnant women with known birth outcomes and haemoglobin measurements during pregnancy were included. Unadjusted and adjusted risk ratios for the association between maternal anaemia and stillbirth were estimated using modified Poisson regression for each cohort and pooled using a random-effects meta-analysis. Kaplan-Meier survival curves were estimated for foetal survival beyond 28 weeks of gestation across categories of anaemia severity. Findings:The median maternal age of the women was 23 years (IQR 21-25). There were 108,982 (52%) women with moderate to severe anaemia (haemoglobin [Hb] <10 g/dL) and 3595 (1.7%) stillbirths. We found that moderate to severe anaemia at any point during pregnancy was associated with a higher risk of stillbirth after 28 weeks of gestation (aRR 1.27, 95% CI 1.07-1.51). The risk ratio for severe anaemia (<7 g/dL) was 3.12 (95% CI 2.46-3.97), and for moderate anaemia it was 1.20 (95% CI 1.02-1.40) at any gestational age, compared to women with no or mild anaemia. We found that stillbirths occurred around the 29th week of gestation among pregnant women with severe anaemia, compared to around 31 weeks of gestation among other women. Interpretation:These findings highlight that maternal anaemia is an important and potentially modifiable risk factor for stillbirth, with severe anaemia associated with early occurrence of stillbirth. Policies aimed at reducing stillbirths should prioritise comprehensive anaemia prevention and treatment strategies among pregnant women. Funding:Indian Council of Medical Research.
Background:Family Planning (FP) is central to reproductive autonomy; however, structural barriers and restrictive gender norms limit equitable access to modern contraceptive methods in South Asia. We aimed to synthesize the current evidence on determinants of modern contraceptive use among married women of reproductive age in South Asia. Methods:We searched Medline, Scopus, Embase, and CINAHL from 1st January, 2015 to 23rd April, 2026. Observational studies from South Asian countries using multivariate analyses to examine determinants of modern contraceptive use among married women of reproductive age (15-49 years) were included. Risk of bias was assessed using the Joanna Briggs Institute (JBI) tool. Generic inverse variance method was used to estimate pooled odds ratios and 95% confidence intervals in RevMan. The review was registered with PROSPERO(CRD42024542764). Findings:There were 78 studies with 2,212,823 study participants that met the inclusion criteria and 70 studies were included in the meta-analysis. Women's formal education and employment (OR: 1.35, 95% CI: 1.08-1.70, I2 = 74; OR: 1.20, 95% CI: 1.01-1.42, I2 = 0%), spousal involvement in decision-making (OR: 2.79, 95% CI: 1.17-6.65, I2 = 66%), access to health facility with FP services (OR: 2.13, 95% CI: 1.12-4.05, I2 = 0%), media exposure (OR: 1.26, 95% CI: 1.10-1.45, I2 = 0%), and counselling from healthcare workers (OR: 1.57, 95% CI: 1.05-2.35, I2 = 0%) were associated with higher modern contraceptive use, but some estimates were based on limited studies hence should be interpreted with caution. Interpretation:Strengthening women's empowerment, increasing male engagement, and improving access of FP services are essential for equitable access of modern contraceptive methods across South Asia. Funding:This study was not funded by any individual or agency.
Background:Women in Nepal commonly experience intimate partner violence. This study aims to understand how abortion access, a key indicator of reproductive autonomy, may be linked to partner violence in this setting. Methods:Between April 2019 and December 2020, pregnant women seeking an abortion at 22 facilities across all provinces in Nepal were recruited into a prospective cohort study. The primary exposure was pregnancy outcome (gave birth vs. did not give birth). The outcome, intimate partner violence, was measured at baseline and then every 12 months using indicators of emotional, physical, economic, and sexual violence. Sociodemographic, health, and care-seeking characteristics were collected at baseline. Mixed effects multivariable logistic regression models, adjusted via propensity score weighting, were used to assess the association between birth and subsequent intimate partner violence via risk differences (RDs). Findings:A total of 1742 pregnant women were followed for a median duration of 62 months. At baseline, 15% reported experiencing emotional violence in the past year, 12% physical, seven percent economic, and nine percent sexual violence. Weighted models showed no significant change in violence risk from baseline among individuals who gave birth (five-year RD = 0.02; 95% CI -0.09 to 0.14). In contrast, overall violence declined among those who did not give birth (one-year RD = -0.05; 95% CI -0.08 to -0.03, five-year RD = -0.08; 95% CI -0.16 to -0.01). This group also experienced declines in physical and economic violence specifically. In comparing the two groups, women who did not give birth experienced less violence overall than those who gave birth, particularly in years two (RD = 0.05; 95% CI 0.00 to 0.10), three (RD = 0.07; 95% CI 0.01 to 0.13), four (RD = 0.07; 95% CI 0.01 to 0.13), and five (RD = 0.10; 95% CI 0.03 to 0.16) after abortion seeking. Interpretation:Compared to those who were able to end their pregnancy, women who gave birth were at higher risk of partner violence, even several years after abortion seeking. Abortion access may offer protection against intimate partner violence compared to carrying an unwanted pregnancy to term. Funding:This study was supported by the National Institutes of Health and the David and Lucile Packard Foundation.
