Prevalence and risk factors of stillbirths among pregnant women from twelve high-volume birthing facilities of Karachi, Pakistan: a longitudinal cohort study

Stillbirths are defined as the death of a fetus at or after 22 weeks of gestation before or during delivery [1]. Despite stillbirths being an important indicator of maternal and child health, the progress achieved in reducing maternal and child mortality has not been paralleled by equivalent advancements in addressing the burden of stillbirths [2]. While the UN Global Strategy for Women’s, Children’s and Adolescents’ Health (2016–2030) and the Every Newborn Action Plan (ENAP) advocate for an end to preventable stillbirths, reducing stillbirths is conspicuously absent as a specific target in the Sustainable Development Goals agenda [3, 4]. This neglect becomes particularly concerning as the burden of stillbirths continues to rise, with approximately 2 million stillborn babies each year, with devastating and far-reaching consequences for families, communities, and health systems [5]. Prevalence of stillbirths is disproportionately higher in low- and middle-income countries (LMICs), where the risk of stillbirth is more than seven times higher than in high-income settings [6]. Over 70% of the global stillbirths occur in Sub-Saharan Africa and South Asia and in 2021, six countries – India, Pakistan, Nigeria, the Democratic Republic of Congo, Ethiopia, and Bangladesh – alone accounted for nearly half of all global stillbirths [5, 7]. With the last standardized global assessment of stillbirth rates conducted in 2019, stillbirths remain underreported or omitted from global data tracking, making it difficult to measure their true burden [2, 5, 7]. If current trends persist, over 16 million additional babies are projected to be stillborn by 2030, with 56 countries likely to miss the ENAP target of reducing stillbirths to 12 or fewer per 1,000 total births [7].

Pakistan is among the 56 countries at risk of missing the ENAP stillbirth target by 2030 [5, 7]. The country also ranks third in the number of stillbirths (> 200,000) and second in stillbirth rates (30.9 per 1000 births) [7]. Pakistan’s progress in reducing rates of stillbirths in the past two decades has been lower (1%) compared to both the regional average (3%) and other low-income countries (1.3%) [7, 8]. Factors contributing to high rates of stillbirths include socioeconomic inequalities, inadequate access to skilled birth attendants and early screening for complications, suboptimal uptake of technological advancements in healthcare and preventable maternal risk factors such as malnutrition, infections, hypertensive disorders, and diabetes [8‐10]. These are compounded by systemic barriers, including under-resourced healthcare systems, weak referral mechanisms, and cultural norms that delay care-seeking behavior [11, 12].

Data on stillbirth prevalence in Pakistan remain fragmented and inconsistent, primarily due to underreporting, the absence of standardized definitions, and infrequent monitoring [8, 12]. The lack of civil and vital registration systems, and comprehensive medical birth and death registries, further complicates the production of reliable and timely stillbirth estimates [2]. Local studies report varying stillbirth rates (6 to 44.4 per 1,000 total births), but these are outdated, limited in scope, typically facility-specific, and fail to capture broader systemic patterns and regional disparities [8, 11, 13, 14]. The impact of geographical disparities, particularly district-level variations and challenges faced by polio-endemic regions, remains underexplored, despite evidence that healthcare access in these areas is often limited. Existing research also often overlooks the need to distinguish stillbirth rates by gestational age, which obscures the differing risk factors and causes of stillbirths at various stages of pregnancy. These gaps in current literature and data systems hinder the identification of key risk factors and the development of effective strategies to reduce stillbirth rates in the country.

We leveraged data from an electronic Pregnant Women and Birth Registry (PWBR) to conduct a facility-based longitudinal cohort research study at selected birthing facilities in Karachi, Pakistan. We estimated the prevalence of stillbirths, analyzed stillbirth rates across different types of birthing facilities, assessed geographic variations, and identified sociodemographic, pregnancy-related, and maternal factors associated with both early and late gestational stillbirths.

Our study evaluated the prevalence and predictors of stillbirths among 13,668 deliveries across 12 birthing facilities in Karachi, Pakistan, revealing an overall weighted stillbirth rate of 12.0 per 1,000 live births with notable variation across facility types and geographic locations. Nearly 50% of stillbirths occurred in preterm deliveries, with the highest weighted stillbirth rate observed in extreme preterm births (< 28 weeks, 429.1 per 1,000 live births). Key maternal health factors, such as ANC visits and lack of TT vaccination were also identified as significant predictors of stillbirth.

Our estimate of the weighted stillbirth rate of 12.0 per 1,000 births is lower than the reported national average of 30.9 per 1,000 total births [7]. We also found substantial variations in the weighted stillbirth rates based on location and facility characteristics. These findings are in line with global and local evidence from other low- and middle-income countries (LMICs) and Pakistan, which report lower than national level estimates for stillbirth rates and highlight variations across geographic locations and facility types, highlighting significant disparities in stillbirth outcomes [7, 12, 28, 29]. This observed deviation from national estimates may stem from several factors, including different study settings (urban vs. rural), underreporting, and challenges in identification of stillbirths by the hospital and healthcare staff, and exclusion of stillbirths occurring at home or enroute to healthcare facilities [12]. Despite this, neither district nor facility type, both assessed at the district level, emerge as significant predictors of stillbirth in our adjusted models. However, at the UC level, we found significantly higher stillbirths in facilities located in the polio-endemic areas (SHRUCs). Higher stillbirth rate in these areas could be because SHRUCs often face compounded challenges, including low maternal literacy, poor socioeconomic conditions, and limited healthcare access and information [28, 30, 31]. This, coupled by the lack of trust people have in the health system, further exacerbates the disparities in maternal and neonatal outcomes [32].

Our study findings revealed the substantial impact of gestational age on stillbirths, with the highest rate observed among extreme preterm deliveries (< 28 weeks) at 429.1 per 1,000 live births, and the lowest among at term deliveries (≥ 37 weeks) at 8.0 per 1,000 live births. This is consistent with findings from other global studies across multiple settings, where preterm delivery is consistently identified as major risk factors for stillbirths [33‐36]. A possible explanation is that preterm birth increases the risk of stillbirth, primarily due to respiratory issues from an immature respiratory system, apnea (prolonged breathing pauses), heart problems, poor temperature regulation caused by insufficient body fat and a weakened immune system that makes infections more likely [34]. The high stillbirth rate among extreme preterm births raises important considerations for clinical care, as these deliveries are often associated with significant neonatal morbidity and mortality. It is worth noting that the stillbirth rate among preterm deliveries nationally may still be an underestimate as the current stillbirth reporting practices in Pakistan, (and many other LMICs), exclude stillbirths occurring before 28 weeks of gestation [14, 37‐40]. Our findings emphasize the need for consistent application of the WHO’s recommended stillbirth definition, which includes stillbirths at any gestational age ≥22 weeks, to ensure more accurate data capture and informed intervention and policy development for timely and high-quality maternal care.

We also found key maternal health related factors which significantly increase the likelihood of stillbirths. Women who had not received TT vaccinations during pregnancy and who had received <8 ANC visits were at higher risk of stillbirth. Literature from other settings has signified the protective effect and safety of maternal immunization during pregnancy for improved birth outcomes and has been recommended by WHO as a key strategy for preventing vaccine-preventable diseases in newborns [14, 37‐39, 41]. Adequate ANC visits (≥8) have been found to be associated with lower stillbirths across multiple settings [42, 43], highlighting the importance of comprehensive and continuous care across pregnancy. ANC contributes to stillbirth prevention by enabling timely detection and management of complications [44], and promoting institutional deliveries attended by skilled birth providers [45].

Household education was also associated with lower stillbirths in our study underscoring the importance of educated family members in improving maternal health and outcomes. Education contributes to improved health outcomes by enhancing communication within families and with healthcare providers, increasing household income and health-seeking capacity, fostering a greater sense of responsibility, and enabling timely recognition and response to danger signs [45, 46].

Our study had certain limitations. First, it was conducted only in selected birthing facilities in Karachi and focused solely on facility-based births (excluding home births) which limited generalizability. Additionally, the accurate identification of stillbirths among health staff was a challenge that has been documented and reported earlier in Pakistan [12]. Due to these challenges, we were unable to collect detailed data on the classification and nature of stillbirths (intrapartum vs. antepartum), which limits the depth of insights we can draw from our results. Additionally, to compute the stillbirth rates in our study, we relied on actual births within the study sample rather than estimated births for the broader setting due to the lack of reliable data. This limitation may affect the generalizability of our findings. Furthermore, we used the ≥ 22-week threshold to define stillbirths, consistent with the WHO/ICD-11 classification. While this enhances sensitivity and aligns with evolving global standards, it may limit comparability with national statistics and prior studies that use the ≥ 28-week threshold. The number of refusals was not recorded due to high patient volume and integration of registration into routine workflows. Based on field team observations, refusals were minimal. Lastly, although approximately one-third of participants were LTF—likely due to delivering outside of study facilities—we conducted a sensitivity analysis using IPW to adjust for differences in baseline characteristics. While this helps mitigate the risk of attrition bias, we acknowledge the potential for residual confounding due to unmeasured variables.

The findings from this study highlight several key areas where policy interventions could significantly reduce stillbirth rates. First, the higher stillbirth rates observed in SHRUC-based and government-operated facilities suggest a need for targeted investments in these settings to improve healthcare infrastructure, staffing, and quality of care. While substantial efforts have been directed toward polio eradication in SHRUCs, our findings underscore the need for a more integrated approach to health interventions. Focusing solely on polio may inadvertently overlook other pressing maternal and child health challenges, including stillbirths. To address this, we recommend that health policies in polio-endemic areas expand their scope to include comprehensive maternal health services. For example, maternal immunizations, such as tetanus toxoid vaccines, should be prioritized alongside routine polio campaigns to reduce infections that may contribute to adverse pregnancy outcomes. Additionally, ANC services must be enhanced in these areas, with a focus on improving both the quality and accessibility of care, a practice that has proven to be successful across multiple countries that have improved maternal and newborn healthcare over the last three decades [47]. Efforts should include better screening for maternal complications, early detection of risk factors, and timely management of conditions that could lead to stillbirths. Leveraging the existing infrastructure of polio programs, such as community outreach, vaccination campaigns, and healthcare worker networks — can facilitate the delivery of integrated maternal and child health services. Finally, given the high prevalence of preterm births and their association with stillbirths, policies that emphasize early identification of preterm labor and provide access to appropriate care and interventions could help reduce the risk of stillbirth, particularly in high-risk pregnancies. Future research is warranted to measure the prevalence of stillbirths at a wider-scale using real-time longitudinal data from electronic birth registries. Investigating the causal relationships between maternal health factors, immunization, and stillbirths, particularly in diverse settings, would offer valuable insights for targeted interventions.

Our study found a stillbirth rate lower than the national average. However, the aggregate number masks underlying disparities between birthing facilities and geographic locations. Polio-endemic regions, in particular, emerged as hotspots for stillbirths, highlighting the need for targeted and integrated interventions. Strengthening maternal health services, through provision of high-quality antenatal care services including maternal immunization and efficient management of medical conditions during pregnancy remains critical to minimize the risk of stillbirths.